There has been some discussion on FWD performance. I thought I would simulate the effects of moving weight forwards or backwards on two FWD cars at Mosport using LapSim to see what the effect is. I did this for an ITR (mine, 1272 kg with stock 162 whp powertrain) and for the ETCC FWD model car supplied with the Bosch program. My base lap time is 1:43.63 or about 8 seconds slower than Pierre Kleinubing's time with a measured 61% front weight bias. The ETCC FWD car (1100 kg with 260 hp) has a 60% base front weight distribution and a 1:25.5 second lap time at Mosport (simulation only) which would be 10 seconds faster than Kleinubing's ITR time. I used 1.08 for max tire grip while the ETCC car has a whopping 1.53 factor for tire grip (slicks rule!). The results were somewhat surprising, in that both cars show the same shape relationship, a parabola, with a minimum at 55% for my ITR and at a slightly different 56.5% for the ETCC car. It would seem that with a FWD, having a 50-50 weight distribution is not the ideal. There are so many old wives' tales to de-bunk. As the % is moved towards the front, the greatest loss in speed occurred in the high speed turn 4, particularly as it starts to go downhill. Less than optimum grip from the rear end I think. Scary stuff when you are lose your rear end in a high speed turn;-)

ITR at Mosport, Laptime vs Front weight %


ETCC FWD at Mosport, Laptime vs Front Weight%


No other parameters were changed except the weight %. Changing spring balance for each weight distribution would likely lead to more insight.

Now you have an idea how much you can improve your laptime by moving your battery to the trunk.

 

good work! for this program how many variables do you have to input? looks like this program could be insanely complex!.

-nate

 

can you isolate FWD or RWD as a variable?

good stuff, i had figured you wanted something like 60/40 for FF, but never had anything substantial to back it up.

 

[QUOTE=tnord]can you isolate FWD or RWD as a variable? [QUOTE]

Yes you can. I tried that. My ITR setup was slower when I changed it to RWD without changing anything else. I did not investigate further as this mod is not in my plans! Maybe someone can try to make a setup file for a Miata and see how it runs.

There are a lot of parameters you can change. In the free version, front weight bias is adjustable, but cross weight is not for example.

 

the program is called lap sim?

 

I always figured that 50/50 was not optimal for a FWD car, it makes sense when you think about it. However, I never had any simulations or real data to back it up. Nice work.

Try upping the HP to 200 or more and I bet you get a bigger %front as the optimal.

 

Remember that this is static weight distribution that we're talking about. It would be interesting to model the same question applying different values for dynamic loads due to F/R aero downforce, that varies with speed.

K

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Knestis &raquo;</TD></TR><TR><TD CLASS="quote">Remember that this is static weight distribution that we're talking about. It would be interesting to model the same question applying different values for dynamic loads due to F/R aero downforce, that varies with speed.

K</TD></TR></TABLE>

I assumed it was using dynamic loads, since it ran the model through a current race track. A static load test like could not be too much of a help, the only program I know of that that can account for dynamic loads and other important parameters is Adams, very powerful stuff.

-nate

 

if you can adjust spring rates, i would think that it attempts to account for dynamic load changes.

*sigh*
it has become painfully obvious that i and everyone else who is somewhat fascinated by this program, truly is a nerd at heart.

 

LapSim is very powerful compared to everyone guessing at what is going on. It is free. It was written up in Race Car Engineering and Race Tech magazines. It uses a very sophisticated Pacjeka tire model, with a 3D grip vs slip angle map. It produces complete charts of all data you would get from a data logger including speed, RPM, g's, gear etc. A lot of people could learn from this thing. What diff/gear set should you use for a track, it will tell you. I just wish I could affford the full version (imagine you can input 216 parameters vs the paltry 26 in the free version). What would be even better is if a bunch of people started using it and made new track maps for their favorite tracks and shared them. You could even have a contest. Set some rules and see who can make the best lap time. A lot more informative than just talking about what street tires to use on your next auto-cross. Might be too much tech though.

The only guy I have ever had dicussions with who actually used the program is on the DSR forum. A lot of those guys use data loggers on their cars. You can see it here: http://b16.ezboard.com/fdsrfor...topic looks like someone is looking for a track file for Portland. Who's gonna make it?


Modified by descartesfool at 2:13 PM 4/23/2004

 

Very impressive, and the results seem to confirm what I thought. [Therefor it must be right, right? ]

As for loosing the back end in turn 4 at Mosport, that would be some MAJOR puckerfactor. I think 4 is scarrier than 2, in corner 2 you just can't see the track or where you are going About as scary as taking a corner middle of forth gear with your eyes closed.

 

i went to a seminar on ADAMS what was impressive. Check it out if you can.

 

how much stock should we put into a free trial version? i only ask because i took the included fwd example and only increased the front bar rate and the car turned a faster lap time at leguna seca. thoughts? it seems to me that the modeling doesn't include all the perculiarities of a fwd setup.

oh, and increasing the rear spring rate or swaybar slowed the car down.

nate

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by MightyMouseTech &raquo;</TD></TR><TR><TD CLASS="quote">Very impressive, and the results seem to confirm what I thought. [Therefor it must be right, right? ]
</TD></TR></TABLE>

Is that what you meant when you said "That is not right at all. 50/50 is pretty much always optimal."

Just wondering.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by rapid_roy &raquo;</TD></TR><TR><TD CLASS="quote">

Is that what you meant when you said "That is not right at all. 50/50 is pretty much always optimal."

Just wondering.</TD></TR></TABLE>

Yes and no??!?! Considering it only adds 0.2 secs/lap, and tire wear will be more even front to back. I always strive to have the car as close to 50/50 as I can get. Which usually is not that close on a FWD car unfortunatley. Although, I have never seen a FWD Honda actually get down to 50/50. It is just what we always aspire too.


Modified by MightyMouseTech at 4:12 PM 4/23/2004

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by solo-x &raquo;</TD></TR><TR><TD CLASS="quote">how much stock should we put into a free trial version? i only ask because i took the included fwd example and only increased the front bar rate and the car turned a faster lap time at leguna seca. thoughts? it seems to me that the modeling doesn't include all the perculiarities of a fwd setup.

oh, and increasing the rear spring rate or swaybar slowed the car down.

nate</TD></TR></TABLE>

That is true. FWD rotation with a stiff rear end may not be very well modelled, since you have to model not only the car but the driver's way of getting the car around the corner. You will notice that the ETCC FWD model they give you has higher front spring (and wheel) rates than the rear. Race Car Engineering article said their software and data were validated on this car and other race cars (total of 5 I think) using actual on board data acquisition. They show an example in the article of actual data compared to simulated data, and it was pretty close. I have no idea if the FWD model has higher front rates in reality. I have never seen setup info on any FWD ETCC car. In any case the model is not very sensitive to wheel rates after the car is set up fairly stiff in roll. I have not found the model to yield great insights into suspension setup, but it is very good at power, weight, gears and tire grip. I use the model to gain insight in the areas where it seems to be clear on what is improving. I was able to get the model to match my lap time without fudging at Mosport within about 1 second (model was faster than me). I used data from my data logger to get parameters for program, such as max tire grip from the lateral accelerometer. I measured drag from coast-down measurements. I thought about contacting the program's author to ask about sensitivity or lack thereof to suspension tuning, but I haven't done it since I thought the support might be a little thin for the free version. Imagine the e-mail they would get! I may try to ask a trick question to get the guy's interest up. Maybe I just need 5300 euros to play the game with the big dogs.

For anyone who wants to read one of the articles: http://www.bosch-motorsport.co...s.pdf


Modified by descartesfool at 6:35 PM 4/23/2004

 

The Archer Brothers built a Renault Le Car way back in the early eighties, and set the chassis up to have almost perfect 50/50 weight distribution. They moved the motor back through the firewall, and jacked other stuff around to get the weight where they wanted it.

The car was undriveable! They could not keep the back end planted whatsoever. The normal chassis with the motor in the right place was much faster.

 

i was just flipping thru an older GRM (not too old, had the white civic Si on the cover), and there was an article from some racing veteran talking about car weights and crossweights and whatnot. he actually says for a FWD, 50/50 is not desirable and in his experience found something like 60/40 or 70/30 better, while for RWD, 50/50 is still clearly the goal and anytihng more all other performance values suffer. we can all have our opinions and simulations or whatever, but im going to trust someone with 40 years of actual racing experience.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Tyson &raquo;</TD></TR><TR><TD CLASS="quote"> we can all have our opinions and simulations or whatever, but im going to trust someone with 40 years of actual racing experience.</TD></TR></TABLE>

The issue is that what works in racing is totally determined by the sanctioning body. If X class was all FWD cars with 50/50 weight bias, and 900 hp.... well, the best car for that class would be 50/50 weight balance, and would do quite well within "racing". Physics has to tell you what theoretically works better than another.

Eh, it's early, and im still awake...grrr

 

great thread