Is this flange too complicated for CNC?
Intake flange, aluminum.
Its about 1" tall. The two points of concern are the runner support colums thats flare out, will the CNC be able to get in there? Its about 27mm between the stacks at the top and 37mm at the base.
Also the interior. It has a contoured transition that goes from the oval shape of the port to the round shape of the runner. Will a CNC be able to get in there and make the shape? Even if it steps it a little thats fine, I can come back and sand it smooth afterwards.
Also one more question. You can see at the top where the runners set in the flange. Is there an epoxy I can use to "cold weld" the runners in place? That way I don't have to worry about warping the flange. The stock manifold I'm modeling this after looks to use a thin epoxy to weld the runners in place, but I have never seen something like that before.



Its about 1" tall. The two points of concern are the runner support colums thats flare out, will the CNC be able to get in there? Its about 27mm between the stacks at the top and 37mm at the base.
Also the interior. It has a contoured transition that goes from the oval shape of the port to the round shape of the runner. Will a CNC be able to get in there and make the shape? Even if it steps it a little thats fine, I can come back and sand it smooth afterwards.
Also one more question. You can see at the top where the runners set in the flange. Is there an epoxy I can use to "cold weld" the runners in place? That way I don't have to worry about warping the flange. The stock manifold I'm modeling this after looks to use a thin epoxy to weld the runners in place, but I have never seen something like that before.



Nothing is too complicated for CNC, but you have to think about the cutting process during design. You're going to CNC mill these, which means spinning round tool.
The only way to cut those external flares would be using a bit which matches that angle and basic length. That type of bit is not readily available, at least not with many angle choices. Is there a reason you can't just make them straight?
The internal bores would be difficult to do. That kind of organic transition can't be cut without stepping for hours. What I would do in this case is machine the circular ports from the top with a tapered end mill, then flip over and machine the oval ports from the bottom with the same tapered end mill. You should be able to draw the intersection of the two tapers as they meet. It won't be perfectly centered, but it will give a similar result.
As far as cold welding goes, bag it. I assume this is a race car part which will stand up to more abuse than some cute epoxy can take. Design your runners to be round tubes with external o-rings and flanges. Design some threaded posts to either side of the ports and slip fit and bolt your runners. If you're going this far, go all the way.
Evan
The only way to cut those external flares would be using a bit which matches that angle and basic length. That type of bit is not readily available, at least not with many angle choices. Is there a reason you can't just make them straight?
The internal bores would be difficult to do. That kind of organic transition can't be cut without stepping for hours. What I would do in this case is machine the circular ports from the top with a tapered end mill, then flip over and machine the oval ports from the bottom with the same tapered end mill. You should be able to draw the intersection of the two tapers as they meet. It won't be perfectly centered, but it will give a similar result.
As far as cold welding goes, bag it. I assume this is a race car part which will stand up to more abuse than some cute epoxy can take. Design your runners to be round tubes with external o-rings and flanges. Design some threaded posts to either side of the ports and slip fit and bolt your runners. If you're going this far, go all the way.
Evan
I agree with your idea of the port transition taking to much time. I redesigned it taking into consideration your idea. Where the two overlap I extended the smaller of the two at a 15* angle to meet up with the other shape. So now its just 90* and 15* angles. What are common angles available in milling? I'd like a little steeper to keep height down, but jumping to 45* would be too steep to air flow IMO. Maybe a 20*.








You'd have to either moddify that outter angled lip to that of an of-the-shelf dovetail, find a place with a TINY dovetail that uses adjustable inserts, or find a place that can custom-grind angles into an exsisting dovetail. The 2 latter will take much more time and money. Given the fact that will have tubular runners welded on, I don't see a need for it - I'd just go with a 90* sidewall.
As for the inside, I think the new design would cut milling time in half, as you could cut half of it with 'one' final pass, the rest with a couple steps and maybe a moddified endmill to save some time.
CNC is like horsepower in this sense: cheap, complicated, fast - pick two
As for the inside, I think the new design would cut milling time in half, as you could cut half of it with 'one' final pass, the rest with a couple steps and maybe a moddified endmill to save some time.
CNC is like horsepower in this sense: cheap, complicated, fast - pick two
I don't think its too bad, the transition on the OD seems pointless, and would likely require a custom saw (which can be $$) or a ball end saw doing step overs (which would mean $$ in runtime)
but the transition on the OD just requires a small step over, which means long tool path, but at the same time high feed rates can be achieved due to nearly zero chip load.
although tbh, if you had sent the original drawing to me in the first place, I probably would've ignored it.
As for angles, look at MSC or dataflute's website. Tapered end mills are quite common, I believe I have some carbide 15* ball end ones, along with various other tapers. It only really gets spendy when you need one that isn't off the shelf
but the transition on the OD just requires a small step over, which means long tool path, but at the same time high feed rates can be achieved due to nearly zero chip load.
although tbh, if you had sent the original drawing to me in the first place, I probably would've ignored it.
As for angles, look at MSC or dataflute's website. Tapered end mills are quite common, I believe I have some carbide 15* ball end ones, along with various other tapers. It only really gets spendy when you need one that isn't off the shelf
It's not to complicaited to produce, but in may not be cost effective. Why not just cut the flange out of 0.500" and make the transition from round to oval in the tubing? Thats what I normally do, you can bend tubing in a vice or if you want it very repeatable use a pair of dies in a vise.
We can cut it for you. If you have a CAD drawing and all we have to do is program it the cost would be dramatically cheaper. If your interested call the shop and speak w/ Dustin and go over the details.
Trending Topics
The easiest way to cut it instead of going through all programming is to digitize the part on a 5 axis Centroid or Prototrak system. Like someone else said though, you will need a special dovetail to do the OD of the flange. But a small ball-nose end mill will do all the internal cutting and just a plain straight end mill will do the rest.
Go here and you can see a port and polish done on video. Pretty impressive.
http://www.centroidcnc.com/
scroll down and you will see the link for the video.
http://www.centroidcnc.com/
scroll down and you will see the link for the video.
Without reading the whole thread, I know that Sleepers Performance in Hamden, CT is already making these flanges on a 3-axis CNC.
I saw a whole batch of them first hand.
I saw a whole batch of them first hand.
i did the same thing with my flange (transition from oval to round on the ports) it does not take more than an hour/side to cut with proper 3d programming. you just need to do 2 setups. it is very basic actually. as for the outsides, just make them straight. there is no need to have a draft on the outside imo.
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by HybridKOOP »</TD></TR><TR><TD CLASS="quote">Without reading the whole thread, I know that Sleepers Performance in Hamden, CT is already making these flanges on a 3-axis CNC.
I saw a whole batch of them first hand.</TD></TR></TABLE>
Interesting since I designed those from my head
They are not a Honda part (nor VW related).
How does a mill do chamfers? Like for a 1/8" (3.175mm) round over on the back of a circle, does the mill use an actual round over bit or does it just step straight lines?
Also when I take these designs to a machinist, what should the file be like? Can the CNC software just read like a CATIA solids design and go from there or do I need to have 2D curves as well? Meaning can it look at a 3D solid and replicate the design straight to the work piece?
I saw a whole batch of them first hand.</TD></TR></TABLE>
Interesting since I designed those from my head
They are not a Honda part (nor VW related).How does a mill do chamfers? Like for a 1/8" (3.175mm) round over on the back of a circle, does the mill use an actual round over bit or does it just step straight lines?
Also when I take these designs to a machinist, what should the file be like? Can the CNC software just read like a CATIA solids design and go from there or do I need to have 2D curves as well? Meaning can it look at a 3D solid and replicate the design straight to the work piece?
Thread
Thread Starter
Forum
Replies
Last Post
covertperformance
Welding / Fabrication
4
Mar 25, 2009 05:56 PM




