LapSim by Bosch - Some amazing results for an ITR at Mosport
01-18-2004, 01:22 PM
#1
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
LapSim by Bosch - Some amazing results for an ITR at Mosport
LapSim software by Bosch is an amazing piece of work. The fact that you can get a fully functioning copy for free that does complete lap simulations is the treat! During the many hours of Dakar coverage this month, I simulated a bunch (as in hundreds) of laps of Mosport in my ITR, and made mods to the car in software to see the effects on the lap time, which is all I care about (well I do prefer passing other cars, but it is even more fun passing the faster ones).
The first thing to do was to see if I could get the software simulator to match my actual lap times. I only drove my ITR one weekend at Mosport this year, and it was a cold October weekend. My car weighs about 2804 lbs with me in it and full of fuel, and it has a completely stock power train. No engine mods whatsoever. Suspension mods include Advance Design coil-overs with 550 lbs front springs and 650 lbs rear, Skunk 2 A-arms, Ingalls rear camber kit, stock sway bars, and Pilot Sport Cup tires, 225/50/15 on 15x7” Team Dynamics Prorace 1 wheels with 45 mm offset. Car is lowered about 1.5” all around, and camber is -2.5 front and -1.7 rear, with toe about 1/16” front and 0” rear. I have a Race Technology DL-90 GPS based data logger, so I know all my speeds anywhere on the track, lap times and lateral and longitudinal g’s (plus a few other things). My best lap time was 1:44.8, with tires that never really warmed up due to the cool temps (7-10 degrees C or 45-50 F). Kleinubing ran about 1:36 in his full race ITR. My car has most of the stock interior except easy out rear stuff and a roll-bar with 5 point belts which adds about 70 lbs + extinguisher and camera.
I used a 162 whp dyno curve as typical for a stock ITR, and I adjusted the tire grip in the software to 1.08 which was the maximum lateral grip from my data logger in slow corners. After inputting all the parameters in the software, it predicted a 1:43.63 lap time vs. my actual lap time of 1:44.8. Stunning!!! I had to add heights to the Mosport track layout supplied with the demo since the demo tracks are flat. I pretty much guessed at the heights from memory. The fact that the predicted and real lap times are so similar was very confidence inspiring. Lap times for the supplied LMP and F3 cars are very close to the real ones (around 1:12 and 1:16). I assumed that I was not pushing my car as fast as it could go, and so the predicted lap time being 1.17 seconds faster than me (on coldish tires) seemed acceptable. Now the fun begins.
First up was increasing whp to see what can be achieved in terms of lap times. LapSim allows you to enter the complete dyno curve every 200 RPM. I did this once and then used its option of raising hp by a percentage, since this was a look and see type of investigation. If hp is increased from 162 to 167, than all points are increased by same %, which in this case is 5hp/162hp or 3.09%. Thus the 100 hp point on the curve goes to 103 hp. I did this for 5 hp steps from 162 to 212 hp and here are the results:
ITR US Gears & US Diff
Mass(kg). Weight (lb).
1272......2804
Power (hp).Lap time (min)
162.......01:43.63
167.......01:43.27
172.......01:42.75
177.......01:42.37
182.......01:42.04
187.......01:41.57
192.......01:41.23
197.......01:40.88
202.......01:40.50
207.......01:40.20
212.......01:39.89
So a 50 whp increase with the same shape of dyno curve gives an improvement in lap time of 3.74 seconds for a 2804 lb ITR. Thus an increase in power of 50/162 whp or 30.9% results in a lap time improvement of 3.74/103.63 seconds or 3.6%. Thus a 10% increase in power should give a 1.16% improvement in lap time, if the result is linear. More about this later.
I then decided to reduce the car’s weight by 45 kg (about 100 lbs) and ran the simulator again. Each lap simulated takes about one minute after you have changed parameters. The results are as follows:
ITR US Gears & US Diff
Mass (kg). Weight (lb).
1227......2705
Power (hp).Lap time (min)
162.......01:43.07
167.......01:42.66
172.......01:42.21
177.......01:41.79
182.......01:41.39
187.......01:41.00
192.......01:40.65
197.......01:40.30
202.......01:39.96
207.......01:39.67
212.......01:39.36
So now we see an improvement in lap time from 1:43.63 to 1:43.07 due to a 99 lb reduction in weight for a 162 hp car, or 0.56 seconds. Car’s weight went down by 99/2804 lbs or 3.5% and lap time went down by 0.56/103.63 seconds or 0.54%. Thus a 10% reduction in weight should correspond to a 1.54% reduction in lap time, compared to a 1.16% improvement for a 10% power increase. I then lowered the weight a further 45 kg and got the following results:
ITR US Gears & US Diff
Mass (kg). Weight (lb).
1182......2606
Power (hp).Lap time (min)
162.......01:42.42
167.......01:42.02
172.......01:41.58
177.......01:41.18
182.......01:40.78
187.......01:40.39
192.......01:40.03
197.......01:39.68
202.......01:39.36
207.......01:39.08
212.......01:38.79
So at this weight, we lower from 2705 lbs to 2606 lbs, and lap time goes from 1:43.07 down to 1:42.42. A 0.65 second reduction corresponds to a 0.65/103.07 seconds or 0.63% reduction for a 99/2705 lb or 3.66% reduction in weight. Thus a 10% reduction in weight should correspond to a 1.78% reduction in lap time, which is slightly more proportionally than before. One last iteration down another 45 kg to 2507 lbs, and we get:
ITR US Gears & US Diff
Mass(kg). Weight (lb).
1137......2507
Power(hp).Lap time(min)
162.......01:41.86
167.......coming soon!
172.......coming soon!
177.......coming soon!
182.......coming soon!
187.......coming soon!
192.......01:39.53
197.......coming soon!
202.......coming soon!
207.......coming soon!
212.......coming soon!
Here we drop 99 lbs and lap time goes to 1:41.86. This is a 0.56 second or .55% time reduction for a 3.8% reduction in weight, giving a 1.45% lap time reduction for a 10 % reduction in weight, in the same range as before. Thus lap time is approximately linearly proportional to weight reduction. Gives you a good formula to equalize racing classes! I found this to be most interesting since I had not seen this written in any book. You also get this same approximately linear relationship to lap time vs. power. I decided to plot some of these data points and let Excel give me the equation for lap time vs. whp for my current ITR setup.
The chart shows a that a slight curve (polynomial) is the form that best matches the lap time to the whp for each weight. One needs a few extra significant digits to use the equations given on the chart (these are from the “add trendline” function in Excel). Each point on the chart is one simulation in LapSim. The resulting equations are:
For 2804 lbs and peak whp and same dyno curve shape I used:
Lap time (sec)= t = .000276*whp^2 -.1756*whp + 125.07
For 2705 lbs:
Lap time (sec)= t = .000342*whp^2 -.2028*whp + 126.94
For 2606 lbs:
Lap time (sec)= t = .000347*whp^2 -.2024*whp + 126.27
And here is a chart sowing the relationship of lap time vs. weight
Here the relationship of lap time vs. weight is almost perfectly linear for the to whp values of 162 and 192 whp. Amazing!
Just figure out how fast you want to go, and now you know where to spend your money! My target is sub 1:40 at Mosport. Looks like I can do it with 193 whp and a 198 lb weight reduction program to 2606 lbs (from the last equation which gives 99.94 sec.), and it might be easier is I change to some new Hoosiers and drive when it is a little warmer. Might not have to lose so many pounds! I will however have to drive a little faster. Curiously, I'm wondering about gears.
I was truly surprised about the very simple relationships between lap time and whp and weight, considering how complex the LapSim model is, with full Pacejka tire model and 3D maps for each tire with slip angle vs. lateral force dependences. I wonder if Carroll Smith knew this. He sure didn't put it into any of his books.
Now we all know a little more than we did.
The first thing to do was to see if I could get the software simulator to match my actual lap times. I only drove my ITR one weekend at Mosport this year, and it was a cold October weekend. My car weighs about 2804 lbs with me in it and full of fuel, and it has a completely stock power train. No engine mods whatsoever. Suspension mods include Advance Design coil-overs with 550 lbs front springs and 650 lbs rear, Skunk 2 A-arms, Ingalls rear camber kit, stock sway bars, and Pilot Sport Cup tires, 225/50/15 on 15x7” Team Dynamics Prorace 1 wheels with 45 mm offset. Car is lowered about 1.5” all around, and camber is -2.5 front and -1.7 rear, with toe about 1/16” front and 0” rear. I have a Race Technology DL-90 GPS based data logger, so I know all my speeds anywhere on the track, lap times and lateral and longitudinal g’s (plus a few other things). My best lap time was 1:44.8, with tires that never really warmed up due to the cool temps (7-10 degrees C or 45-50 F). Kleinubing ran about 1:36 in his full race ITR. My car has most of the stock interior except easy out rear stuff and a roll-bar with 5 point belts which adds about 70 lbs + extinguisher and camera.
I used a 162 whp dyno curve as typical for a stock ITR, and I adjusted the tire grip in the software to 1.08 which was the maximum lateral grip from my data logger in slow corners. After inputting all the parameters in the software, it predicted a 1:43.63 lap time vs. my actual lap time of 1:44.8. Stunning!!! I had to add heights to the Mosport track layout supplied with the demo since the demo tracks are flat. I pretty much guessed at the heights from memory. The fact that the predicted and real lap times are so similar was very confidence inspiring. Lap times for the supplied LMP and F3 cars are very close to the real ones (around 1:12 and 1:16). I assumed that I was not pushing my car as fast as it could go, and so the predicted lap time being 1.17 seconds faster than me (on coldish tires) seemed acceptable. Now the fun begins.
First up was increasing whp to see what can be achieved in terms of lap times. LapSim allows you to enter the complete dyno curve every 200 RPM. I did this once and then used its option of raising hp by a percentage, since this was a look and see type of investigation. If hp is increased from 162 to 167, than all points are increased by same %, which in this case is 5hp/162hp or 3.09%. Thus the 100 hp point on the curve goes to 103 hp. I did this for 5 hp steps from 162 to 212 hp and here are the results:
ITR US Gears & US Diff
Mass(kg). Weight (lb).
1272......2804
Power (hp).Lap time (min)
162.......01:43.63
167.......01:43.27
172.......01:42.75
177.......01:42.37
182.......01:42.04
187.......01:41.57
192.......01:41.23
197.......01:40.88
202.......01:40.50
207.......01:40.20
212.......01:39.89
So a 50 whp increase with the same shape of dyno curve gives an improvement in lap time of 3.74 seconds for a 2804 lb ITR. Thus an increase in power of 50/162 whp or 30.9% results in a lap time improvement of 3.74/103.63 seconds or 3.6%. Thus a 10% increase in power should give a 1.16% improvement in lap time, if the result is linear. More about this later.
I then decided to reduce the car’s weight by 45 kg (about 100 lbs) and ran the simulator again. Each lap simulated takes about one minute after you have changed parameters. The results are as follows:
ITR US Gears & US Diff
Mass (kg). Weight (lb).
1227......2705
Power (hp).Lap time (min)
162.......01:43.07
167.......01:42.66
172.......01:42.21
177.......01:41.79
182.......01:41.39
187.......01:41.00
192.......01:40.65
197.......01:40.30
202.......01:39.96
207.......01:39.67
212.......01:39.36
So now we see an improvement in lap time from 1:43.63 to 1:43.07 due to a 99 lb reduction in weight for a 162 hp car, or 0.56 seconds. Car’s weight went down by 99/2804 lbs or 3.5% and lap time went down by 0.56/103.63 seconds or 0.54%. Thus a 10% reduction in weight should correspond to a 1.54% reduction in lap time, compared to a 1.16% improvement for a 10% power increase. I then lowered the weight a further 45 kg and got the following results:
ITR US Gears & US Diff
Mass (kg). Weight (lb).
1182......2606
Power (hp).Lap time (min)
162.......01:42.42
167.......01:42.02
172.......01:41.58
177.......01:41.18
182.......01:40.78
187.......01:40.39
192.......01:40.03
197.......01:39.68
202.......01:39.36
207.......01:39.08
212.......01:38.79
So at this weight, we lower from 2705 lbs to 2606 lbs, and lap time goes from 1:43.07 down to 1:42.42. A 0.65 second reduction corresponds to a 0.65/103.07 seconds or 0.63% reduction for a 99/2705 lb or 3.66% reduction in weight. Thus a 10% reduction in weight should correspond to a 1.78% reduction in lap time, which is slightly more proportionally than before. One last iteration down another 45 kg to 2507 lbs, and we get:
ITR US Gears & US Diff
Mass(kg). Weight (lb).
1137......2507
Power(hp).Lap time(min)
162.......01:41.86
167.......coming soon!
172.......coming soon!
177.......coming soon!
182.......coming soon!
187.......coming soon!
192.......01:39.53
197.......coming soon!
202.......coming soon!
207.......coming soon!
212.......coming soon!
Here we drop 99 lbs and lap time goes to 1:41.86. This is a 0.56 second or .55% time reduction for a 3.8% reduction in weight, giving a 1.45% lap time reduction for a 10 % reduction in weight, in the same range as before. Thus lap time is approximately linearly proportional to weight reduction. Gives you a good formula to equalize racing classes! I found this to be most interesting since I had not seen this written in any book. You also get this same approximately linear relationship to lap time vs. power. I decided to plot some of these data points and let Excel give me the equation for lap time vs. whp for my current ITR setup.
The chart shows a that a slight curve (polynomial) is the form that best matches the lap time to the whp for each weight. One needs a few extra significant digits to use the equations given on the chart (these are from the “add trendline” function in Excel). Each point on the chart is one simulation in LapSim. The resulting equations are:
For 2804 lbs and peak whp and same dyno curve shape I used:
Lap time (sec)= t = .000276*whp^2 -.1756*whp + 125.07
For 2705 lbs:
Lap time (sec)= t = .000342*whp^2 -.2028*whp + 126.94
For 2606 lbs:
Lap time (sec)= t = .000347*whp^2 -.2024*whp + 126.27
And here is a chart sowing the relationship of lap time vs. weight
Here the relationship of lap time vs. weight is almost perfectly linear for the to whp values of 162 and 192 whp. Amazing!
Just figure out how fast you want to go, and now you know where to spend your money! My target is sub 1:40 at Mosport. Looks like I can do it with 193 whp and a 198 lb weight reduction program to 2606 lbs (from the last equation which gives 99.94 sec.), and it might be easier is I change to some new Hoosiers and drive when it is a little warmer. Might not have to lose so many pounds! I will however have to drive a little faster. Curiously, I'm wondering about gears.
I was truly surprised about the very simple relationships between lap time and whp and weight, considering how complex the LapSim model is, with full Pacejka tire model and 3D maps for each tire with slip angle vs. lateral force dependences. I wonder if Carroll Smith knew this. He sure didn't put it into any of his books.
Now we all know a little more than we did.
01-18-2004, 03:11 PM
#3
Honda-Tech Member
Join Date: Feb 2000
Location: Snowwhitepillowformybigfathead
Posts: 4,049
Likes: 0
Received 2 Likes
on
2 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
Thanks Claude.
So you grew up to be one of those kids who does their homework while you watch TV eh?
Scott, who always suspected that with more horsepower and/or less weight one might theoretically go faster - " I KNEW it!"
So you grew up to be one of those kids who does their homework while you watch TV eh?
Scott, who always suspected that with more horsepower and/or less weight one might theoretically go faster - " I KNEW it!"
01-18-2004, 10:24 PM
#4
I really don't know
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (RR98ITR)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by RR98ITR »</TD></TR><TR><TD CLASS="quote">Thanks Claude.
So you grew up to be one of those kids who does their homework while you watch TV eh?
Scott, who always suspected that with more horsepower and/or less weight one might theoretically go faster - " I KNEW it!"</TD></TR></TABLE>
hmmm so thats how it works....
Good write up on all that stuff. I'd love to play around with that data type stuff. I love statistics and numbers. I played with a datalogger preview software for hours just looking at different numbers throughout the track and where/how much the person braked etc. etc...
So you grew up to be one of those kids who does their homework while you watch TV eh?
Scott, who always suspected that with more horsepower and/or less weight one might theoretically go faster - " I KNEW it!"</TD></TR></TABLE>
hmmm so thats how it works....
Good write up on all that stuff. I'd love to play around with that data type stuff. I love statistics and numbers. I played with a datalogger preview software for hours just looking at different numbers throughout the track and where/how much the person braked etc. etc...
01-19-2004, 12:40 AM
#6
Join Date: Jan 2004
Location: Southern California, california, usa
Posts: 16
Likes: 0
Received 0 Likes
on
0 Posts
Re: (RacerZook)
Holy crap!
That is absolutely brilliant. I've road raced for years and have always assigned those two values (wieght/horsepower) to mods i've done to anticipate how much quicker i'd go on race day.
I've always subscribed to the " 100lb. deduction = a 10hp increase. Your scientific
tests shows that it would take closer to 200lbs. to equal 10hp. While i'm generalizing...is this close to accurate?
Awesome information.
01-19-2004, 02:31 AM
#7
Honda-Tech Member
Join Date: Jul 2003
Location: Delaware, OH, 43015
Posts: 542
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by descartesfool »</TD></TR><TR><TD CLASS="quote">LapSim software by Bosch is an amazing piece of work. The fact that you can get a fully functioning copy for free that does complete lap simulations is the treat! </TD></TR></TABLE>
Free? Where?
Free? Where?
Trending Topics
01-19-2004, 05:05 AM
#8
New User
Join Date: Aug 2001
Location: Durham, NC, USA
Posts: 591
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (jpl95si)
All this work to gain more power and/or lose weight and I bet you can come close to your goals by getting stickier tires.
Did you run the simulation with higher G ratings? Peak G's in the neighborhood of 1.2 to 1.3 ought to be feasible on Hoosiers in warm weather. That should translate into much better corner speeds, exit speeds and shorter braking distances. All of which will help more than a little extra power (since power only helps straight line acceleration).
Scott
Did you run the simulation with higher G ratings? Peak G's in the neighborhood of 1.2 to 1.3 ought to be feasible on Hoosiers in warm weather. That should translate into much better corner speeds, exit speeds and shorter braking distances. All of which will help more than a little extra power (since power only helps straight line acceleration).
Scott
01-19-2004, 08:22 AM
#9
Trial User
Join Date: Jan 2000
Location: Plotting My Revenge
Posts: 7,722
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (RR98ITR)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by RR98ITR »</TD></TR><TR><TD CLASS="quote">Scott, who always suspected that with more horsepower and/or less weight one might theoretically go faster - " I KNEW it!"</TD></TR></TABLE>
Really?
No ****?
Damn. All these years I've been tryin to cheat by stuffing lead putty in the nooks and crannys of my car. Figuring that, like a bobsled, more weight would make me faster in downhill sections. And that pizza and Chunky Monkey diet has been leading me in the wrong direction.
Damn damn damn.
Really?
No ****?
Damn. All these years I've been tryin to cheat by stuffing lead putty in the nooks and crannys of my car. Figuring that, like a bobsled, more weight would make me faster in downhill sections. And that pizza and Chunky Monkey diet has been leading me in the wrong direction.
Damn damn damn.
01-19-2004, 08:25 AM
#10
Honda-Tech Member
Join Date: Jan 2002
Location: 13.9 at 101.76mph, 2420lbs, FL, USA
Posts: 5,690
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
The problem with this is you are assuming a linear relationship between lap times and hp increase. While this is alright for your experimentals it has no application in the real world. If you every plot the hp vs. 1/4 times you will understand this very quickly. Just for conjecture I'll throw out some numbers, all things being equal. Let's say a 100 whp 2500lb car can do 16sec in 1/4. Add 50 whp, it's down to 15secs. Add 50whp more and it drops to 14.5. Add 50 more and your at 14.25. Now, I'm just running these numbers off the top of my head, but if you get some actuals you'll notice the relationship between your lap times and hp is an exponential relationship, not a linear one.
Unless I'm a total idiot and then I just wasted the last 3 minutes of my life.
By the way, that software program sounds freaking awesome.
Unless I'm a total idiot and then I just wasted the last 3 minutes of my life.
By the way, that software program sounds freaking awesome.
01-19-2004, 09:46 AM
#11
Join Date: Oct 2003
Location: Pittsburgh, Pa, USA
Posts: 294
Likes: 0
Received 0 Likes
on
0 Posts
Re: (integraracer)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by integraracer »</TD></TR><TR><TD CLASS="quote">
I've always subscribed to the " 100lb. deduction = a 10hp increase. Your scientific
tests shows that it would take closer to 200lbs. to equal 10hp. While i'm generalizing...is this close to accurate?
</TD></TR></TABLE>
i heard its closer to 6 lbs/ hp. but, i dont know **** - i drive a porsche
also, is any of the weight in your graphs unsprung?
I've always subscribed to the " 100lb. deduction = a 10hp increase. Your scientific
tests shows that it would take closer to 200lbs. to equal 10hp. While i'm generalizing...is this close to accurate?
</TD></TR></TABLE>
i heard its closer to 6 lbs/ hp. but, i dont know **** - i drive a porsche
also, is any of the weight in your graphs unsprung?
01-19-2004, 02:15 PM
#12
Honda-Tech Member
Join Date: Feb 2000
Posts: 906
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (jpl95si)
01-19-2004, 03:20 PM
#13
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
Having found a relationship between lap time and whp and weight at Mosport for my ITR, I thought I would look into gears and differential ratios. The results are valid for the shape of power curve I used and would be different for curves with the power peak at different RPM's. Since there was a little more Dakar coverage, I had some more time to ponder. The power curve was input at every 200 RPM, so if you change the location of the power peak, you will change lap time due to whp and its linked relationship to the gearing. You can input any complete dyno curve you want into the software.
There are three common gearsets for the ITR, and 3 common differentials. They are the USDM, the JDM and the ATS setups. This gives 3x3 or 9 combinations to try.
The gear sets are as follows:
Gearset.....1.........2.........3..........4...... ..5
ga(stock)..3.230...2.105...1.458...1.107...0.848
gb(JDM)....3.230...2.105...1.458...1.034...0.787
gc(ATS)....3.077...2.105...1.652...1.308...1.033
and the diffs are
Da(stock)..4.400
Db(JDM)....4.785
Dc(ATS)....4.928
The LapSim software requires you to enter the number of gear teeth for each pair, so I just worked out some hypothetical pairings that match the above ratios. For example, a 4.400 diff can be represented by a 66:15 ratio, a 4.785 works out to about a 14:67 ratio and a 4.928 is 14:69. There have always been a lot of opinions about which setup is the fastest. In WC, the diff and gears are mandated, so it is not an issue. But if you have the option, which is fastest. With my stock power setup, I never get into 5 th gear down (really up) the main straight at Mosport. So what is the effect of gears and diff for various power levels and weights. The results were a bit of a surprise to me, and cerainly more confidence inspiring than ll the opinions I have read:
Weight (lb).Power (hp).Gear-Diff combo
.................US-US.....US-JDM....US-ATS...JDM-US....JDM-JDM..JDM-ATS..ATS-US...ATS-JDM..ATS-ATS
2804..162..01:43.62..01:43.30..01:43.20..01:44.10. .01:43.78..01:43.63..01:42.96..01:43.07..01:43.06
2804..172..01:42.73..01:42.48..01:42.39..01:43.19. .01:42.92..01:42.79..01:42.15..01:42.28..01:42.33
2804..182..01:41.92..01:41.71..01:41.66..01:42.33. .01:42.12..01:42.03..01:41.40..01:41.56..01:41.71
2804..192..01:41.16..01:41.02..01:40.94..01:41.54. .01:41.40..01:41.33..01:40.65..01:40.95..01:41.12
2705..162..01:43.15..01:42.87..01:42.78..01:43.65. .01:43.35..01:43.20..01:42.55..01:42.64..01:42.64
2705..172..01:42.25..01:42.02..01:41.96..01:42.70. .01:42.45..01:42.35..01:41.72..01:41.84..01:41.93
2705..182..01:41.43..01:41.25..01:41.22..01:41.84. .01:41.68..01:41.59..01:41.00..01:41.15..01:41.29
2705..192..01:40.69..01:40.53..01:40.49..01:41.06. .01:40.93..01:40.86..01:40.21..01:40.52..01:40.73
It is obvious that changing the differential from US to JDM to ATS improves the lap time. For the 162 whp case at 2804 lbs, laptime goes from 103.62 seconds down to 103.30 and to 103.20 for the JDM and ATS diffs respectively. That is an improvement of .32 and .42 seconds. Very significant, and I think Michael Schumacher would kill for that kind of advantage. But looking down the table, the JDM-JDM combo is slower, surprise, surprise. What is fast is the ATS close ratio gear set combined with the US diff: 102.96 seconds. That is a jaw-dropping 0.66 second lap time improvement! Now Paul Stoddard is having a look.
From the first equation of laptime vs. whp (for the 2804 lb case), it takes 170 whp to get the same laptime improvement. Thus the ATS-US combo is equivalent to 8 whp for the 2804 lb case! When the whp is increased and weight decreased, the gains are less "monumental". On the last line of the table, for the US gearset, the time goes down from 100.69 to 100.53 and to 100.49 for the US, JDM and ATS diffs respectively. Smaller gains of 0.16 and 0.20 seconds over stock. Again, the ATS-US combo proves to be fastest at 102.21 with a gain of 0.48 seconds. You're moving up the grid, and smiling. Your competitors can't figure out how you got that half second: cheating bastard! It turns out that for these examples, the best choice by far is the US diff with an ATS gear set. It gives the lowest lap time for every case. I never would have thought! And here I was about to buy a new set of diff gears. The worst case? The overall loser is the JDM gearset with US diff. It has the worst lap time in every case. As for the last two cases, the ATS-JDM and ATS-ATS, the top speed is maxed out on the straight, at 190.8 and 185.5 km/hr respectively. The max speed for the US-US is 236 km/hr while for the ATS-US it is 207.6 km/hr.
The data in chart form for the 2804 lb case:
The data in chart form for the 2705 lb case:
It leads me to believe that someone at ATS knew what they were doing. And what does this bliss cost you ask? Looks like $1320 for the close ratio 1-3-4-5 gearset from ATS plus a tranny rebuild.
http://www.a-t-s-usa.com/ats-p...shtml
And the ATS diff set? That would be $785.
http://www.a-t-s-usa.com/ats-p...shtml
Gear set only is the obvious better choice here, and only $535 more than the diff.
Speed costs. How fast do you want to go?
P.S. You are gonna have to shift a whole lot more around the track using the fast ATS-US combo. From the Bosch LapSim results graphs, the US-US combo has a total of 10 shifts for one lap while the ATS-US winner combo needs 16 shifts total, and that assumes 160 ms shifts for both. The first combo never makes into 5 th gear, while the latter has 4 places in 5th around the track. But of course skill is not lacking, only money eh!
Modified by descartesfool at 7:43 PM 1/19/2004
There are three common gearsets for the ITR, and 3 common differentials. They are the USDM, the JDM and the ATS setups. This gives 3x3 or 9 combinations to try.
The gear sets are as follows:
Gearset.....1.........2.........3..........4...... ..5
ga(stock)..3.230...2.105...1.458...1.107...0.848
gb(JDM)....3.230...2.105...1.458...1.034...0.787
gc(ATS)....3.077...2.105...1.652...1.308...1.033
and the diffs are
Da(stock)..4.400
Db(JDM)....4.785
Dc(ATS)....4.928
The LapSim software requires you to enter the number of gear teeth for each pair, so I just worked out some hypothetical pairings that match the above ratios. For example, a 4.400 diff can be represented by a 66:15 ratio, a 4.785 works out to about a 14:67 ratio and a 4.928 is 14:69. There have always been a lot of opinions about which setup is the fastest. In WC, the diff and gears are mandated, so it is not an issue. But if you have the option, which is fastest. With my stock power setup, I never get into 5 th gear down (really up) the main straight at Mosport. So what is the effect of gears and diff for various power levels and weights. The results were a bit of a surprise to me, and cerainly more confidence inspiring than ll the opinions I have read:
Weight (lb).Power (hp).Gear-Diff combo
.................US-US.....US-JDM....US-ATS...JDM-US....JDM-JDM..JDM-ATS..ATS-US...ATS-JDM..ATS-ATS
2804..162..01:43.62..01:43.30..01:43.20..01:44.10. .01:43.78..01:43.63..01:42.96..01:43.07..01:43.06
2804..172..01:42.73..01:42.48..01:42.39..01:43.19. .01:42.92..01:42.79..01:42.15..01:42.28..01:42.33
2804..182..01:41.92..01:41.71..01:41.66..01:42.33. .01:42.12..01:42.03..01:41.40..01:41.56..01:41.71
2804..192..01:41.16..01:41.02..01:40.94..01:41.54. .01:41.40..01:41.33..01:40.65..01:40.95..01:41.12
2705..162..01:43.15..01:42.87..01:42.78..01:43.65. .01:43.35..01:43.20..01:42.55..01:42.64..01:42.64
2705..172..01:42.25..01:42.02..01:41.96..01:42.70. .01:42.45..01:42.35..01:41.72..01:41.84..01:41.93
2705..182..01:41.43..01:41.25..01:41.22..01:41.84. .01:41.68..01:41.59..01:41.00..01:41.15..01:41.29
2705..192..01:40.69..01:40.53..01:40.49..01:41.06. .01:40.93..01:40.86..01:40.21..01:40.52..01:40.73
It is obvious that changing the differential from US to JDM to ATS improves the lap time. For the 162 whp case at 2804 lbs, laptime goes from 103.62 seconds down to 103.30 and to 103.20 for the JDM and ATS diffs respectively. That is an improvement of .32 and .42 seconds. Very significant, and I think Michael Schumacher would kill for that kind of advantage. But looking down the table, the JDM-JDM combo is slower, surprise, surprise. What is fast is the ATS close ratio gear set combined with the US diff: 102.96 seconds. That is a jaw-dropping 0.66 second lap time improvement! Now Paul Stoddard is having a look.
From the first equation of laptime vs. whp (for the 2804 lb case), it takes 170 whp to get the same laptime improvement. Thus the ATS-US combo is equivalent to 8 whp for the 2804 lb case! When the whp is increased and weight decreased, the gains are less "monumental". On the last line of the table, for the US gearset, the time goes down from 100.69 to 100.53 and to 100.49 for the US, JDM and ATS diffs respectively. Smaller gains of 0.16 and 0.20 seconds over stock. Again, the ATS-US combo proves to be fastest at 102.21 with a gain of 0.48 seconds. You're moving up the grid, and smiling. Your competitors can't figure out how you got that half second: cheating bastard! It turns out that for these examples, the best choice by far is the US diff with an ATS gear set. It gives the lowest lap time for every case. I never would have thought! And here I was about to buy a new set of diff gears. The worst case? The overall loser is the JDM gearset with US diff. It has the worst lap time in every case. As for the last two cases, the ATS-JDM and ATS-ATS, the top speed is maxed out on the straight, at 190.8 and 185.5 km/hr respectively. The max speed for the US-US is 236 km/hr while for the ATS-US it is 207.6 km/hr.
The data in chart form for the 2804 lb case:
The data in chart form for the 2705 lb case:
It leads me to believe that someone at ATS knew what they were doing. And what does this bliss cost you ask? Looks like $1320 for the close ratio 1-3-4-5 gearset from ATS plus a tranny rebuild.
http://www.a-t-s-usa.com/ats-p...shtml
And the ATS diff set? That would be $785.
http://www.a-t-s-usa.com/ats-p...shtml
Gear set only is the obvious better choice here, and only $535 more than the diff.
Speed costs. How fast do you want to go?
P.S. You are gonna have to shift a whole lot more around the track using the fast ATS-US combo. From the Bosch LapSim results graphs, the US-US combo has a total of 10 shifts for one lap while the ATS-US winner combo needs 16 shifts total, and that assumes 160 ms shifts for both. The first combo never makes into 5 th gear, while the latter has 4 places in 5th around the track. But of course skill is not lacking, only money eh!
Modified by descartesfool at 7:43 PM 1/19/2004
01-19-2004, 05:18 PM
#14
New User
Join Date: Aug 2001
Location: Durham, NC, USA
Posts: 591
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
160 ms shifts, eh? Try factoring it in with a realistic shift time like 500 ms (or do you have a sequential gearbox?).
I'd still like to see the effect of a little more grip. All this work to see what gearing and pwoer do when some stiffer springs and stickier tires will do more (should be easy to confirm that speculation).
Scott
I'd still like to see the effect of a little more grip. All this work to see what gearing and pwoer do when some stiffer springs and stickier tires will do more (should be easy to confirm that speculation).
Scott
01-19-2004, 05:52 PM
#15
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (celica73)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by celica73 »</TD></TR><TR><TD CLASS="quote">160 ms shifts, eh? Try factoring it in with a realistic shift time like 500 ms (or do you have a sequential gearbox?).
I'd still like to see the effect of a little more grip. All this work to see what gearing and pwoer do when some stiffer springs and stickier tires will do more (should be easy to confirm that speculation).
Scott</TD></TR></TABLE>
You can simulate anything you want. Just do it and post your results. If I changed shift times to 500 ms, nothing would be learned. I already know all my real shift times from my data logger. If I change the shift time to 260 ms fom 160 ms, the lap times for the US-US case go up by 0.12 seconds, while the ATS-US case goes up by 0.18 seconds due to ther being three more upshifts per lap, so we can reduce gain fom ATS-US combo by 0.06 seconds. I say this is of no consequence here. However asuming you shift at one particlar speed, this just adds a constant to the lap time. I want to know want improves lap time. Also, just put grip to 1.6 g's for your tires and your lap times will be great. I can also add a ton of downforce to the car and lap times improve. Run a couple of hundred simulations and tell us what you learned. This analysis of variations has taught me a ton of things I did not know, and that are not described in any books I have seen. I did however suspect that a lighter car with more power was better, and that matching gearing to the engine's power band and track was a good idea. But now I can figure out $/second of lap time reduction.
Modified by descartesfool at 7:28 AM 1/20/2004
I'd still like to see the effect of a little more grip. All this work to see what gearing and pwoer do when some stiffer springs and stickier tires will do more (should be easy to confirm that speculation).
Scott</TD></TR></TABLE>
You can simulate anything you want. Just do it and post your results. If I changed shift times to 500 ms, nothing would be learned. I already know all my real shift times from my data logger. If I change the shift time to 260 ms fom 160 ms, the lap times for the US-US case go up by 0.12 seconds, while the ATS-US case goes up by 0.18 seconds due to ther being three more upshifts per lap, so we can reduce gain fom ATS-US combo by 0.06 seconds. I say this is of no consequence here. However asuming you shift at one particlar speed, this just adds a constant to the lap time. I want to know want improves lap time. Also, just put grip to 1.6 g's for your tires and your lap times will be great. I can also add a ton of downforce to the car and lap times improve. Run a couple of hundred simulations and tell us what you learned. This analysis of variations has taught me a ton of things I did not know, and that are not described in any books I have seen. I did however suspect that a lighter car with more power was better, and that matching gearing to the engine's power band and track was a good idea. But now I can figure out $/second of lap time reduction.
Modified by descartesfool at 7:28 AM 1/20/2004
01-20-2004, 08:11 AM
#16
Honda-Tech Member
Join Date: Apr 2002
Location: toronto, Canada
Posts: 23
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by descartesfool »</TD></TR><TR><TD CLASS="quote"> Kleinubing ran about 1:36 in his full race ITR. </TD></TR></TABLE>
FYI, PK ran about 1:32 with BFG-R1, and he ran 1:33.2 with T1-S (with lower front-spoiler and big rear wing) in 2002.
Modified by B2 at 9:26 AM 1/20/2004
FYI, PK ran about 1:32 with BFG-R1, and he ran 1:33.2 with T1-S (with lower front-spoiler and big rear wing) in 2002.
Modified by B2 at 9:26 AM 1/20/2004
01-20-2004, 09:40 AM
#17
Honda-Tech Member
Join Date: Jan 2002
Location: 13.9 at 101.76mph, 2420lbs, FL, USA
Posts: 5,690
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by descartesfool »</TD></TR><TR><TD CLASS="quote">
The LapSim software requires you to enter the number of gear teeth for each pair, so I just worked out some hypothetical pairings that match the above ratios. For example, a 4.400 diff can be represented by a 66:15 ratio, a 4.785 works out to about a 14:67 ratio and a 4.928 is 14:69. </TD></TR></TABLE>
How do you figure this out?
The LapSim software requires you to enter the number of gear teeth for each pair, so I just worked out some hypothetical pairings that match the above ratios. For example, a 4.400 diff can be represented by a 66:15 ratio, a 4.785 works out to about a 14:67 ratio and a 4.928 is 14:69. </TD></TR></TABLE>
How do you figure this out?
01-20-2004, 10:37 AM
#18
Honda-Tech Member
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (DarkKnight[DMD])
gear ratio = ratio of teeth. just find the lowest common denominator.
4.4:1
66/4.4 = 15
as far as the shift time - if your datalogger says thats how long it takes, then good job, but two points:
I'd say that this would be a VERY important data point to be dead-on with... also, what are you using to detirmine shift time? acceleration resuming, or throttle being re-applied?
not trying shoot holes in your theories, as it seems very concrete.
4.4:1
66/4.4 = 15
as far as the shift time - if your datalogger says thats how long it takes, then good job, but two points:
I'd say that this would be a VERY important data point to be dead-on with... also, what are you using to detirmine shift time? acceleration resuming, or throttle being re-applied?
not trying shoot holes in your theories, as it seems very concrete.
01-21-2004, 01:53 AM
#19
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (B2)
[QUOTE=B2]
FYI, PK ran about 1:32 with BFG-R1, and he ran 1:33.2 with T1-S (with lower front-spoiler and big rear wing) in 2002.
[QUOTE]
Your information is incorrect. He never got close to the times you mention. Pierre Kleinubing ran a race best lap of 1:35.461 at Mosport in 2002 in his ITR, see:
http://www.speedvisionwc.com/2....html
In practice, he ran a best lap of 1:36.166 and in qualifying he ran a best time of 1:35.611, slower than his best race lap, see:
http://www.speedvisionwc.com/2002/mos/tc-grid.html
In 2001, PK ran a best race lap of 1:35.574 at Mosport, see:
http://www.speedvisionwc.com/2....html
In 2000, PK ran a best race lap of 1:34.715 at Mosport on BFG-R1 tires, see:
http://www.world-challenge.com/00/tc-results.html
Modified by descartesfool at 7:32 AM 1/21/2004
FYI, PK ran about 1:32 with BFG-R1, and he ran 1:33.2 with T1-S (with lower front-spoiler and big rear wing) in 2002.
[QUOTE]
Your information is incorrect. He never got close to the times you mention. Pierre Kleinubing ran a race best lap of 1:35.461 at Mosport in 2002 in his ITR, see:
http://www.speedvisionwc.com/2....html
In practice, he ran a best lap of 1:36.166 and in qualifying he ran a best time of 1:35.611, slower than his best race lap, see:
http://www.speedvisionwc.com/2002/mos/tc-grid.html
In 2001, PK ran a best race lap of 1:35.574 at Mosport, see:
http://www.speedvisionwc.com/2....html
In 2000, PK ran a best race lap of 1:34.715 at Mosport on BFG-R1 tires, see:
http://www.world-challenge.com/00/tc-results.html
Modified by descartesfool at 7:32 AM 1/21/2004
01-21-2004, 08:55 AM
#21
Honda-Tech Member
Join Date: Jan 2002
Location: 13.9 at 101.76mph, 2420lbs, FL, USA
Posts: 5,690
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (Greyout)
well ok... I'm still not sure exactly what you mean...soo....
could you convert these gears into those numbers?
1. 3.307
2. 1.950
3. 1.360
4. 1.071
5. .87
FD 4.266
could you convert these gears into those numbers?
1. 3.307
2. 1.950
3. 1.360
4. 1.071
5. .87
FD 4.266
01-22-2004, 02:38 AM
#22
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
For the next installment in this LapSim series, we have drag racing Euro style.
Another nice aspect of LapSim is that it simulates the famous standing kilometer, which was always the way that Porsche used to compare cars. None of this wimpy short track ¼ mile stuff. While giving you the time and speed at the end of 1000m, it also gives you times for 100m and 400m. The 400m distance is the metric equivalent of the ¼ mile, which converts to 402m. The software plots two curves, one for longitudinal acceleration which shows shift points and another plot of speed in km/hr vs. time. You can c ick on the speed graph and get the time to accelerate to any speed.
While my interest in drag racing is severely limited, I am still curious about how much the straight line times vary with weight, whp and gearing. So another few episodes of SpeedTV (Victory by Design, or "I'm Alain de Cadenet and you're not"), and I've got the answers. The plots also list the time to speed for 0-100 km/hr (62.1 mph) and 0-160 km/hr (99.4 mph).By the way the software is metric only. If you require a nice little free unit converter, you can get one here:
http://www.esbconsult.com
And now the results. I started with my current weight of 2804 lbs and power of 162 whp, and I varied the gearset-diff:
Weight (lb).Power (hp)........
2804........162
Times to speed and distance, and speeds in km/hr
Gear-Diff....US-US.US-JDM.US-ATS.JDM-US.JDM-JDM.JDM-ATS.ATS-US.ATS-JDM.ATS-ATS
0_100kph.......6.94.....7.22....7.24......6.94.... ..7.22......7.24......7.00.....6.98......6.96
0_160kph.....17.66....17.22...17.08....18.22....17 .22...17.82...16.78....16.98....16.94
100m............6.60......6.54....6.54......6.60.. ....6.54.....6.54.....6.68......6.52.....6.52
400m..........15.20....15.14...15.16....15.20....1 5.14....15.18...15.18....15.00...15.00
1000m........27.64....27.58...27.60....27.84....27 .58....27.68...27.54....27.35...27.40
kph@1000.193.00..190.00..190.00..191.00..191.00..1 90.00..194.00..191.00..186.00
kph-max...236.00..230.80..226.50..234.00..235.40..234. 10..207.60..190.80..185.50
The times are remarkably close to the times reported in the original 1997 Car & Driver test, which had a 1/4 mile time of 15.3 seconds where LapSim gives 15.20 seconds for 400m, and Car & Driver got a 0_100 mph time of 17.9 seconds while the model here gives 17.66 seconds for 0_99 mph (my test weight, whp, shift times and max longitudinal tire slip angle could have been different than the Car & Driver test).
Now the results are a little different than for the lap times in terms of selecting optimum gearset-diff ratio. Fastest 0-62 mph is the US-US setup. But check out the 0-99 mph. Fastest is ATS-US, by almost a full second over the US-US case. Quarter mile time (400m) is best with ATS-JDM or ATS-ATS, by 0.20 seconds. 1000m time is best with the ATS-JDM at 27.35 seconds, but highest speed at 1000m is 194 km/hr for the ATS-US case. Remember that the ATS-US case gave all the best lap times at Mosport. It is not obvious which gearset-diff combo gives the best lap time by looking at the straight line acceleration results. Since I know the ATS-US combo is best for lap times, should I look at 0-160kph time and speed at 1000m? Not sure at this point.
So how do times improve with different power and weight. I got distracted by Alain's discourse (or is that diatribe?) and didn't do every combo. But the one's I did show what improvement can be expected for the US-US and the ATS-US combos.
Now use same weight, but put power up by 30 (we now add 0-200kph time since car can boogie!)
Weight (lb).Power (hp)........
2804........192
Times to speed and distance, and speeds in km/hr
Gear-Diff.....US-US..ATS-US
0-100kph.......6.24....6.16
0-160kph.....14.90...14.12
0-200kph.....25.64...24.10
100m.............6.40....6.36
400m..........14.56...14.32
1000m........26.30...26.08
kph@1000.201.00..205.00
kph-max...246.80..207.60
A new revelation. The ATS-US combo beats the US-US combo in every case, except top speed, which is RPM limited to 207.6 km/hr due to the ATS 5th gear ratio. Stock gearing gives a nice top speed of almost 247 km/hr or 153 mph. You're going to need a long straight though.
Now we go back to stock power but test weight drops to 2705 lbs (no 200 klics time):
Weight (lb).Power (hp)........
2705........162
Times to speed and distance, and speeds in km/hr
Gear-Diff....US-US..ATS-US
0-100kph........6.74....6.76
0-160kph.....17.08...16.20
100m.............6.52....6.58
400m..........15.04...14.98
1000m........27.36...27.24
kph@1000.193.00..194.00
kph-max...236.60..207.60
The winner is not so clear-cut here, but I like the 0-160 time for the ATS-US combo, 16.20 seconds compared to 17.08 seconds. Compared to the stock weight case, the ¼ mile time goes down from 15.20 to 15.04 seconds for the US gearset and from 15.18 to 14.98 for the ATS-US setup. Going to need to diet a little more to get those fast times.
Now low weight, power up by 30 (we now get back our 0-200kph time)
Weight (lb).Power (hp)........
2705........192
Times to speed and distance, and speeds in km/hr
Gear-Diff..US-US..ATS-US
0-100kph.......6.16....6.08
0-160kph....14.52...13.76
0-200kph....24.86...23.38
100m............6.38....6.36
400m..........14.48...14.32
1000m........26.12...25.88
kph@1000.203.00..207.00
kph-max...247.40..207.60
I still like the ATS-US combo. Sub 14 second time to 160 kph and sub 24 second time to 200 kph. You may notice that the top speed is increased for the US-US case compared to the 2804 lb/192 whp case by 0.6 km/hr. Must be the lower rolling resistance of the tires in the model. Subtle!
So if you are a little more interested in acceleration times than me, you can just figure out what is the perfect combo for your own car. You can adjust the launch RPM and tire grip to see the effect.
By the way, someone mentioned that you can get higher g’s than I used in the simulations (1.08 g’s in cold temps) by changing tires. I got 1.3 g’s peak at Watkins Glen this summer, but that is on banked turns. You have to use the g’s you get on a flat slow turn for the model. I never got 1.3 g’s on a flat turn, but close to 1.2 is possible. So many parameters and so little time! Maybe in the next installment.
Another nice aspect of LapSim is that it simulates the famous standing kilometer, which was always the way that Porsche used to compare cars. None of this wimpy short track ¼ mile stuff. While giving you the time and speed at the end of 1000m, it also gives you times for 100m and 400m. The 400m distance is the metric equivalent of the ¼ mile, which converts to 402m. The software plots two curves, one for longitudinal acceleration which shows shift points and another plot of speed in km/hr vs. time. You can c ick on the speed graph and get the time to accelerate to any speed.
While my interest in drag racing is severely limited, I am still curious about how much the straight line times vary with weight, whp and gearing. So another few episodes of SpeedTV (Victory by Design, or "I'm Alain de Cadenet and you're not"), and I've got the answers. The plots also list the time to speed for 0-100 km/hr (62.1 mph) and 0-160 km/hr (99.4 mph).By the way the software is metric only. If you require a nice little free unit converter, you can get one here:
http://www.esbconsult.com
And now the results. I started with my current weight of 2804 lbs and power of 162 whp, and I varied the gearset-diff:
Weight (lb).Power (hp)........
2804........162
Times to speed and distance, and speeds in km/hr
Gear-Diff....US-US.US-JDM.US-ATS.JDM-US.JDM-JDM.JDM-ATS.ATS-US.ATS-JDM.ATS-ATS
0_100kph.......6.94.....7.22....7.24......6.94.... ..7.22......7.24......7.00.....6.98......6.96
0_160kph.....17.66....17.22...17.08....18.22....17 .22...17.82...16.78....16.98....16.94
100m............6.60......6.54....6.54......6.60.. ....6.54.....6.54.....6.68......6.52.....6.52
400m..........15.20....15.14...15.16....15.20....1 5.14....15.18...15.18....15.00...15.00
1000m........27.64....27.58...27.60....27.84....27 .58....27.68...27.54....27.35...27.40
kph@1000.193.00..190.00..190.00..191.00..191.00..1 90.00..194.00..191.00..186.00
kph-max...236.00..230.80..226.50..234.00..235.40..234. 10..207.60..190.80..185.50
The times are remarkably close to the times reported in the original 1997 Car & Driver test, which had a 1/4 mile time of 15.3 seconds where LapSim gives 15.20 seconds for 400m, and Car & Driver got a 0_100 mph time of 17.9 seconds while the model here gives 17.66 seconds for 0_99 mph (my test weight, whp, shift times and max longitudinal tire slip angle could have been different than the Car & Driver test).
Now the results are a little different than for the lap times in terms of selecting optimum gearset-diff ratio. Fastest 0-62 mph is the US-US setup. But check out the 0-99 mph. Fastest is ATS-US, by almost a full second over the US-US case. Quarter mile time (400m) is best with ATS-JDM or ATS-ATS, by 0.20 seconds. 1000m time is best with the ATS-JDM at 27.35 seconds, but highest speed at 1000m is 194 km/hr for the ATS-US case. Remember that the ATS-US case gave all the best lap times at Mosport. It is not obvious which gearset-diff combo gives the best lap time by looking at the straight line acceleration results. Since I know the ATS-US combo is best for lap times, should I look at 0-160kph time and speed at 1000m? Not sure at this point.
So how do times improve with different power and weight. I got distracted by Alain's discourse (or is that diatribe?) and didn't do every combo. But the one's I did show what improvement can be expected for the US-US and the ATS-US combos.
Now use same weight, but put power up by 30 (we now add 0-200kph time since car can boogie!)
Weight (lb).Power (hp)........
2804........192
Times to speed and distance, and speeds in km/hr
Gear-Diff.....US-US..ATS-US
0-100kph.......6.24....6.16
0-160kph.....14.90...14.12
0-200kph.....25.64...24.10
100m.............6.40....6.36
400m..........14.56...14.32
1000m........26.30...26.08
kph@1000.201.00..205.00
kph-max...246.80..207.60
A new revelation. The ATS-US combo beats the US-US combo in every case, except top speed, which is RPM limited to 207.6 km/hr due to the ATS 5th gear ratio. Stock gearing gives a nice top speed of almost 247 km/hr or 153 mph. You're going to need a long straight though.
Now we go back to stock power but test weight drops to 2705 lbs (no 200 klics time):
Weight (lb).Power (hp)........
2705........162
Times to speed and distance, and speeds in km/hr
Gear-Diff....US-US..ATS-US
0-100kph........6.74....6.76
0-160kph.....17.08...16.20
100m.............6.52....6.58
400m..........15.04...14.98
1000m........27.36...27.24
kph@1000.193.00..194.00
kph-max...236.60..207.60
The winner is not so clear-cut here, but I like the 0-160 time for the ATS-US combo, 16.20 seconds compared to 17.08 seconds. Compared to the stock weight case, the ¼ mile time goes down from 15.20 to 15.04 seconds for the US gearset and from 15.18 to 14.98 for the ATS-US setup. Going to need to diet a little more to get those fast times.
Now low weight, power up by 30 (we now get back our 0-200kph time)
Weight (lb).Power (hp)........
2705........192
Times to speed and distance, and speeds in km/hr
Gear-Diff..US-US..ATS-US
0-100kph.......6.16....6.08
0-160kph....14.52...13.76
0-200kph....24.86...23.38
100m............6.38....6.36
400m..........14.48...14.32
1000m........26.12...25.88
kph@1000.203.00..207.00
kph-max...247.40..207.60
I still like the ATS-US combo. Sub 14 second time to 160 kph and sub 24 second time to 200 kph. You may notice that the top speed is increased for the US-US case compared to the 2804 lb/192 whp case by 0.6 km/hr. Must be the lower rolling resistance of the tires in the model. Subtle!
So if you are a little more interested in acceleration times than me, you can just figure out what is the perfect combo for your own car. You can adjust the launch RPM and tire grip to see the effect.
By the way, someone mentioned that you can get higher g’s than I used in the simulations (1.08 g’s in cold temps) by changing tires. I got 1.3 g’s peak at Watkins Glen this summer, but that is on banked turns. You have to use the g’s you get on a flat slow turn for the model. I never got 1.3 g’s on a flat turn, but close to 1.2 is possible. So many parameters and so little time! Maybe in the next installment.
01-27-2004, 05:25 PM
#23
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
One element which affects the lap time significantly is the value assigned to the tire grip parameter. In my basic simulations, I used the value of 1.08 which is the value in lateral g's I got from my data logger for low speed corners, where it is easy to get the tires to maximum g's, and down force or lift does not have any effect.
The software allows a change to both the front and rear values separately such that different wear and temperature effects can be modeled. I changed the front and rear equally since I had no data to indicate otherwise, and I am using the same size tire (225/50/15) at all four corners.
Here is a sample of lap time variation for the base case I started with:
Weight (lb).Power (hp).Gear/diff.
2804.........162............US-US
Grip..........1.08......1.10.......1.12.......1.14 ......1.16......1.18.......1.20......1.22
Laptime.103.62...103.16...102.70...102.46...102.04 ...101.61...101.23...100.78
The main point which can be seen is that the lap tie is very sensitive to the value assigned to tire grip. If we compare to the previous simulations for whp and weight starting with the base case, it takes a whp increase from 162 to 192 to exactly match the laptime improvement for a grip change from 1.08 to 1.20.
Here is a chart with an equation to predict lap time change vs. grip:
The chart shows the relationship to be very close to linear over the sample range. Surprising!!
I have repeated grip data from above and from the first post to make the comparison more obvious.
ITR US Gears & US Diff
Mass (kg). Weight (lb)
1272......2804..............162 whp for all grip changes
Power (hp).Lap time (min).......Grip (g's).Lap time
162............01:43.63..............1.08.......01 :43.63
167............01:43.27..............1.10.......01 :43.16
172............01:42.75..............1.12.......01 :42.70
177............01:42.37..............1.14.......01 :42.46
182............01:42.04..............1.16.......01 :42.04
187............01:41.57..............1.18.......01 :41.61
192............01:41.23..............1.20.......01 :41.23
197............01:40.88..............1.22.......01 :40.78
202............01:40.50..............
207............01:40.20..............
212............01:39.89..............
This match is only a coincidence as it will change when the mass of the car changes, but it is interesting to see that the relationship of lap time to grip is close to being linear. If grip is changed by 10%, say from 1.10 to 1.21, lap time improves by about 103.16-101.00 or 2.16 seconds which works out to 2.1%. From the first post, a 10% increase in power should give a 1.16% improvement in lap time, or about half of what you get for the same % increase in tire grip. Tire wars anyone? And re-capping the effect of weight changes, a 10% reduction in weight should correspond to a 1.54% reduction in lap time.
Comparing the effects:
10% increase in whp = 1.16% reduction in lap time
10% decrease in mass= 1.54% reduction in lap time
10% increase in grip= 2.16% reduction in lap time
These effects are from my simulations for an ITR at Mosport starting with 162 whp and 2804 lb test weight, but I think they would be similar for other cars and tracks. I have learned a few more things I did not know. Now if I could just figure out how to add $ signs to the equations.
Modified by descartesfool at 7:14 AM 6/5/2004
The software allows a change to both the front and rear values separately such that different wear and temperature effects can be modeled. I changed the front and rear equally since I had no data to indicate otherwise, and I am using the same size tire (225/50/15) at all four corners.
Here is a sample of lap time variation for the base case I started with:
Weight (lb).Power (hp).Gear/diff.
2804.........162............US-US
Grip..........1.08......1.10.......1.12.......1.14 ......1.16......1.18.......1.20......1.22
Laptime.103.62...103.16...102.70...102.46...102.04 ...101.61...101.23...100.78
The main point which can be seen is that the lap tie is very sensitive to the value assigned to tire grip. If we compare to the previous simulations for whp and weight starting with the base case, it takes a whp increase from 162 to 192 to exactly match the laptime improvement for a grip change from 1.08 to 1.20.
Here is a chart with an equation to predict lap time change vs. grip:
The chart shows the relationship to be very close to linear over the sample range. Surprising!!
I have repeated grip data from above and from the first post to make the comparison more obvious.
ITR US Gears & US Diff
Mass (kg). Weight (lb)
1272......2804..............162 whp for all grip changes
Power (hp).Lap time (min).......Grip (g's).Lap time
162............01:43.63..............1.08.......01 :43.63
167............01:43.27..............1.10.......01 :43.16
172............01:42.75..............1.12.......01 :42.70
177............01:42.37..............1.14.......01 :42.46
182............01:42.04..............1.16.......01 :42.04
187............01:41.57..............1.18.......01 :41.61
192............01:41.23..............1.20.......01 :41.23
197............01:40.88..............1.22.......01 :40.78
202............01:40.50..............
207............01:40.20..............
212............01:39.89..............
This match is only a coincidence as it will change when the mass of the car changes, but it is interesting to see that the relationship of lap time to grip is close to being linear. If grip is changed by 10%, say from 1.10 to 1.21, lap time improves by about 103.16-101.00 or 2.16 seconds which works out to 2.1%. From the first post, a 10% increase in power should give a 1.16% improvement in lap time, or about half of what you get for the same % increase in tire grip. Tire wars anyone? And re-capping the effect of weight changes, a 10% reduction in weight should correspond to a 1.54% reduction in lap time.
Comparing the effects:
10% increase in whp = 1.16% reduction in lap time
10% decrease in mass= 1.54% reduction in lap time
10% increase in grip= 2.16% reduction in lap time
These effects are from my simulations for an ITR at Mosport starting with 162 whp and 2804 lb test weight, but I think they would be similar for other cars and tracks. I have learned a few more things I did not know. Now if I could just figure out how to add $ signs to the equations.
Modified by descartesfool at 7:14 AM 6/5/2004
02-01-2004, 06:27 PM
#24
Honda-Tech Member
Thread Starter
Join Date: Jan 2003
Location: Cogito ergo sum, Canada
Posts: 1,979
Likes: 0
Received 0 Likes
on
0 Posts
Re: LapSim by Bosch - Some amazing results for an ITR at Mosport (descartesfool)
The Aero effect.
24 hours of Daytona. All those cars have big wings. And it rained and rained. Red flag drag. So what better time to look at the influence of downforce on the lap times for my ITR at Mosport with the help of LapSim.
I have never seen published numbers for the front and rear lift/downforce coefficients for a type R, so in all my previous simulations I set both the front and rear coefficients to 0. I had previously measured the drag coefficient with coast-downs using my data logger, and using a frontal area of 1.90 m2, the drag coefficient turns out to be 0.33. These are the values I used for all the other simulations. As a point of interest, the software demo gives 4 sample cars, a Le Mans Prototype (like an Audi R8), a Formula 3 car, a European Touring Car Championship FWD car, and a "Sports Car" representing a high powered road car. Here are the values for the supplied cars to give a reference point, along with tire grip. The base values I used are listed as well.
Parameter....Area...Drag..F Down.R Down.F Grip.R Grip
LMP..............1.70...0.50...0.50.....0.75.....1 .60...1.70
F3................1.50...0.40...0.55.....1.10..... 1.45...1.55
ETCC FWD....1.80...0.33...0.20.....0.10.....1.53...1.53
Sports Car....1.81...0.31...0.00.....0.00.....1.03...1.09
ITR..............1.90...0.33...0.00.....0.00.....1 .08...1.08
The very high grip for the LMP, F3 and ETCC FWD show the cars are on slicks, with low speed g values of 1.5 or more. The Sports Car and my ITR have much lower values, closer to 1.1 g’s. The model predicts the lateral acceleration around the whole lap, and for the cars with the most downforce, it shows the lateral g’s are much higher in the faster corners. In the fast corners #1, #4 and #8 (the one after the main straight) the g value I well above 20 m/s2, while in the slower corners (# 9 & 10) following the one after the main straight, the g values are just shy of 20 m/2 in #9 and around 18 m/s2 in #10 which is the slowest of the last 3 corners. Predicted speeds are 200, 150 and 130 km/hr in those 3 corners. It must be exciting to drive a real race car around there where you corner with more g’s the faster the corner is. It would take some adjustment;-)
The two purpose built race cars have high drag and very high downforce coefficients, up to a staggering 1.10 for the Formula 3 rear wing. The ETCC car has coefficients of 0.1 and 0.2, which is what I have assumed could be achieved with a god aero package on a Type R, such as a good rear wing and a front spoiler/splitter combo. Lacking a wind tunnel, I am guessing here, but I think it would be possible to match those Alfas which is the car I suppose is a sample for the LapSim model. (All their cars are referred to in the Racecar Engineering article by the program's author as having been modeled by the program to compare with real car data).
I used the baseline car with 162 hp and 2804 lb weight and increased the front and rear downforce (called lift in the program) coefficients to see how much gain there is. I did not change the drag coefficient to keep things simple, and since I have no clue how much drag is added in real terms when changing the downforce coefficients. Here are the laptime results in seconds:
............Front Downforce Coefficient
Rear.....0.00......0.02.....0.04......0.06.....0.0 8......0.10
0.00..103.63..103.57..103.55..103.53..103.52..103. 49
0.02..103.50..103.48..103.46..103.43..103.41..103. 39
0.04..103.43..103.39..103.35..103.33..103.31..103. 28
0.06..103.45..103.31..103.28..103.25..103.20..103. 16
0.08..103.29..103.23..103.18..103.15..103.10..103. 08
0.10..103.26..103.18..103.12..103.08..103.02..102. 99
0.12..103.22..103.15..103.08..103.02..102.96..102. 93
0.14..103.19..103.09..103.04..102.97..102.93..103. 06
0.16..103.16..103.09..103.01..102.95..102.88..103. 03
0.18..103.12..103.05..102.98..102.91..102.84..102. 99
0.20..103.09..103.00..102.93..102.85..102.79..102. 94
I was surprised that the lap time is not very sensitive to downforce. The biggest improvement was at a front coefficient of 0.08 and a rear value of 0.20. This reduced the lap time from 103.63 seconds to 102.79 seconds, or just 0.84 seconds. I plotted the results in a 3D map to show that there are no sudden effects as downforce is added.
It would appear money is better spent on improving horsepower and tires, as it only took a 10 whp increase on the base case to get the same improvement in lap time (actually 0.88 second improvement). It took a 143 lb reduction to get the same benefit starting from the base case. A tire grip increase from 1.08 to 1.12 gave a 0.93 second better time, and a change to the ATS gearset gave a 0.66 second lower time. As I said my tire grip value was low due to cold outside and track temperatures, but I used the measured value for that cold October day at Mosport. I am expecting a staggering improvement this summer.
All this is helping me to focus my money. Every little bit helps, but some bits are better than others!
24 hours of Daytona. All those cars have big wings. And it rained and rained. Red flag drag. So what better time to look at the influence of downforce on the lap times for my ITR at Mosport with the help of LapSim.
I have never seen published numbers for the front and rear lift/downforce coefficients for a type R, so in all my previous simulations I set both the front and rear coefficients to 0. I had previously measured the drag coefficient with coast-downs using my data logger, and using a frontal area of 1.90 m2, the drag coefficient turns out to be 0.33. These are the values I used for all the other simulations. As a point of interest, the software demo gives 4 sample cars, a Le Mans Prototype (like an Audi R8), a Formula 3 car, a European Touring Car Championship FWD car, and a "Sports Car" representing a high powered road car. Here are the values for the supplied cars to give a reference point, along with tire grip. The base values I used are listed as well.
Parameter....Area...Drag..F Down.R Down.F Grip.R Grip
LMP..............1.70...0.50...0.50.....0.75.....1 .60...1.70
F3................1.50...0.40...0.55.....1.10..... 1.45...1.55
ETCC FWD....1.80...0.33...0.20.....0.10.....1.53...1.53
Sports Car....1.81...0.31...0.00.....0.00.....1.03...1.09
ITR..............1.90...0.33...0.00.....0.00.....1 .08...1.08
The very high grip for the LMP, F3 and ETCC FWD show the cars are on slicks, with low speed g values of 1.5 or more. The Sports Car and my ITR have much lower values, closer to 1.1 g’s. The model predicts the lateral acceleration around the whole lap, and for the cars with the most downforce, it shows the lateral g’s are much higher in the faster corners. In the fast corners #1, #4 and #8 (the one after the main straight) the g value I well above 20 m/s2, while in the slower corners (# 9 & 10) following the one after the main straight, the g values are just shy of 20 m/2 in #9 and around 18 m/s2 in #10 which is the slowest of the last 3 corners. Predicted speeds are 200, 150 and 130 km/hr in those 3 corners. It must be exciting to drive a real race car around there where you corner with more g’s the faster the corner is. It would take some adjustment;-)
The two purpose built race cars have high drag and very high downforce coefficients, up to a staggering 1.10 for the Formula 3 rear wing. The ETCC car has coefficients of 0.1 and 0.2, which is what I have assumed could be achieved with a god aero package on a Type R, such as a good rear wing and a front spoiler/splitter combo. Lacking a wind tunnel, I am guessing here, but I think it would be possible to match those Alfas which is the car I suppose is a sample for the LapSim model. (All their cars are referred to in the Racecar Engineering article by the program's author as having been modeled by the program to compare with real car data).
I used the baseline car with 162 hp and 2804 lb weight and increased the front and rear downforce (called lift in the program) coefficients to see how much gain there is. I did not change the drag coefficient to keep things simple, and since I have no clue how much drag is added in real terms when changing the downforce coefficients. Here are the laptime results in seconds:
............Front Downforce Coefficient
Rear.....0.00......0.02.....0.04......0.06.....0.0 8......0.10
0.00..103.63..103.57..103.55..103.53..103.52..103. 49
0.02..103.50..103.48..103.46..103.43..103.41..103. 39
0.04..103.43..103.39..103.35..103.33..103.31..103. 28
0.06..103.45..103.31..103.28..103.25..103.20..103. 16
0.08..103.29..103.23..103.18..103.15..103.10..103. 08
0.10..103.26..103.18..103.12..103.08..103.02..102. 99
0.12..103.22..103.15..103.08..103.02..102.96..102. 93
0.14..103.19..103.09..103.04..102.97..102.93..103. 06
0.16..103.16..103.09..103.01..102.95..102.88..103. 03
0.18..103.12..103.05..102.98..102.91..102.84..102. 99
0.20..103.09..103.00..102.93..102.85..102.79..102. 94
I was surprised that the lap time is not very sensitive to downforce. The biggest improvement was at a front coefficient of 0.08 and a rear value of 0.20. This reduced the lap time from 103.63 seconds to 102.79 seconds, or just 0.84 seconds. I plotted the results in a 3D map to show that there are no sudden effects as downforce is added.
It would appear money is better spent on improving horsepower and tires, as it only took a 10 whp increase on the base case to get the same improvement in lap time (actually 0.88 second improvement). It took a 143 lb reduction to get the same benefit starting from the base case. A tire grip increase from 1.08 to 1.12 gave a 0.93 second better time, and a change to the ATS gearset gave a 0.66 second lower time. As I said my tire grip value was low due to cold outside and track temperatures, but I used the measured value for that cold October day at Mosport. I am expecting a staggering improvement this summer.
All this is helping me to focus my money. Every little bit helps, but some bits are better than others!
