F430 Header Cat Thermal Expansion and Exhaust Supports

As a new F430 spider owner I have read with great interest and some fear about the dreaded header failures. The solutions seem to lean heavily towards aftermarket headers, removing the trans mounted clamps and switching to Capristo brackets. However, I did not want the additional loudness and the significant costs. As an engineer I wanted some data. Since I found that the PO had used crap spring bolts for the header to cat connection I decided to fix this and get my data while I was there. I have MKII headers with original cats, muffler and tips. After fixing the spring bolts and removing the lower trans mounted clamps (which clearly are a problem) I put the exhaust back together loosely and then snugged up the spring bolts. I then removed the muffler and tips so I just head the unsupported cats and headers. Next I carefully measured the position of the outboard end of the cats and then warmed up the car. To run the car I started it thru the window so as not to move the chassis and mess up my measurements. I let the car run for about 15 minutes and remeasured everything. Here are the results.

Before the warmup, the back end of the cat was at 60F (measured with a thermocouple). At the end of the warmup, the same point was at 325F. All directions are based on me standing at the back of the car looking forward. I measured both sides independently. All dimensions are in inches. As a result of the warm up,

The back end of the right cat moved: up 0.10”, right (towards the outside) 0.03”, back 0.30”.

The back end of the left cat moved: up 0.05”, left 0.0”, back 0.25”.

I am not sure why the left side moved less. But that is for another day.

So what kind of support should the exhaust system have based on thermal expansion alone? Certainly it should have compliance in the fore-aft direction. It should support a little vertical motion. Left to right motion is constrained by the muffler itself so this is not something the support can address. The Capristo mount certainly addresses the fore-aft motion but it appears that the vertical support is reduced as the exhaust moves up because of the opposed springs. Having taught mechanical engineering for 8 years I thought a quick analysis of the original Ferrari “diving board” supports under the muffler might be helpful. I will spare the details but based on the dimensions, material properties and end constraints, the original Ferrari design theoretically requires 31 pounds of force for every tenth of an inch deflection. For the test I ran, if the rear support was connected it would push forward on the header/cat with about 90 pounds when everything heated up. The vertical support would be a similar story but maybe worse.

One thought I had was to reduce the 0.125” thickness of the “diving boards”. If they are replaced with a plate that was 0.060” then the forward force would be reduced to 3.5 pounds for every tenth of an inch deflection. This would be a cheap change but it still does not address the vertical change. If the diving board was tipped so the top was forward and the bottom was back so the plane of the board was tipped 10 degrees forward at the top then the rearward expansion of the header/cat would also cause the support to raise about 1/6 of the rearward displacement. Things are looking better.

I probably have missed some issues but I thought this might get people thinking. Let me know your thoughts. I apologize if you are tired of hearing about headers and mounting but this is all new for me. Cheers.

 

Hi mike32,

Thanks for the new data point. It got me thinking. The thermal expansion of 304 stainless steel is somewhere around 0.00001 in/in-F. The distance from the center of the header to the end of the cat is about 35 inches. It is 37 inches to the midline of the muffler and 44 inches to where the Capristo racket mounts. From my test I had a temperature rise at the end of the cat of about 260 F. So IF the unconstrained header and cat had a constant 260F rise then theoretically the end of the cat would move 260 x 35 x 0.00001 or about 0.09 inches. My axial measurement was from the end of the cat to the end of the stock muffler mount (the diving board) and I saw almost three times that so either something else moved during the test or the average temperature rise of the system was 3 x 260 or 780F. Plus the 60F starting temperature puts me at a temperature of 840F. I do not know what an expected exhaust gas temperature is but maybe someone can chime in.

If I look at your expansion of 18 mm that is 0.71 inches. The distance from the midline of the header to where your mount connects to the car is about 44 inches. So if all the motion you saw came from thermal expansion then theoretically the average temperature rise of your exhaust system including the bracket was 0.71 divided by (0.00001 x 44) or 1614 F. Guessing that you started at 60F then your total exhaust system average temp was 1674 F. That is pretty hot. Just as a data point, 304 stainless is annealed at about 1900F and melts at about 2500F. Again, I do not have a feel for typical exhaust system temps but would be interested if someone else has that info.

I suspect that I am missing some other component causing the rearward displacement. I will try to take some other measurements and will post what I find.

Thanks again for the info.

 

Hi All,

Mike32:
309 stainless expands about 10% less than 304. So the 304 numbers I used are probably a worst case.

I completely agree with your reasoning for the springs on the header-cat connection. The header-cat connection also is somewhat like a ball and socket joint so it allows there to be a "kink" in the axial line thru the two components.

CoreyNJ:
I think the lower trans brackets were just wrong. Certainly the exhaust components are going to expand to the rear as they get hot. To put a rigid constraint "locking" the expanding exhaust to the differently expanding engine-gearbox makes no sense to me. When I took my lower brackets off there was a popping sound because the bracket was trying to squash the exhaust even in the cold state. Whoever put mine together made a bad design worse by how they bolted it up.

I am hesitant to believe that the donut at the header-cat joint expands mush against the springs. I may try to measure that. I am at a bit of a loss to explain why mike32 and I see the magnitude of expansion that we observed.

I wish the 430 exhaust had a flex pipe. It might have saved many a cracked header.

tstuli:
I agree with the you on the merits of the split exhaust sport exhaust. There support design is not very elegant but it certainly appears to do the job.

To all:
Thanks for all your thoughts. I will continue to run some tests and have already made some test modifications to my original exhaust mounts. I will let you know what I find.

 

Well I finally got to test my first exhaust support concept. In putting it on I came upon a realization that the order that you tighten bolts makes a big difference. Initially I had everything a bit loose and then started from the header and worked to the back. The problem is that when you tighten up the header to cat bolts you basically lock up that joint. Then when you try to raise up the muffler to center the tips in the body openings you are basically bending the header/cat assembly. You can do it but you are putting unwanted stress into all of the parts. I think the goal should be to reduce any stress you put into the system. The order I ended up with was
1-tighten the header to the engine.
2-put all the components together but do not tighten any of the bolts. I snugged the header to cat bolts to seat the gasket but then loosened them up several turns
3-tighten the muffler to tips bolts on both side of the muffler.
4-tighten the tip to cat clamps
5-place a 3/4" thick board between the bottom of the tips (both sides) and the opening in the body. This centers the tips pretty well. It also rotates the tip/muffler/cat assembly about the cat/header joint. If the cat-header bolts are too tight then you will have to lift pretty hard to get the tips centered.
6-Now everything is where it should be and the cat-header joint will have rotated to allow both the header and cat to be stress free. Tighten everything else up.

With the system now pretty much stress free I could play around with the muffler support. As many have done, I got rid of the lower trans to cat bracket. It way over constrains the system and can only cause trouble when things heat up. When I was swapping out the original rear support I blocked the tips in a centered position to carry the weight until I had my new bracket in place. Worth noting, the original muffler weighs about 30 pounds and the tips weigh about 10 pounds on each side. For my first try, I replaced the thick long wing of the original system with a 0.06" thick stainless wing on each side.

Image Unavailable, Please Login

Image Unavailable, Please Login

I also got rid of the top plate on the center bracket. In my view, this center bracket is a bizarre design. The bracket is super stiff but it is mounted with three bolts that are very close to each other. The strength of the bracket way exceeds the strength of the mounting. When I bolted my dual thin wings up they went together with no bending or muscling (the tips were still supported in the body openings). After everything was snug I removed the boards centering the tips. The tips stayed centered. I yanked on the system many ways and it felt very secure. This design provides a rigid vertical support but is pretty soft in the front to back direction. The freed up center bracket also can open up to allow the muffler to grow as it heats up.

I once again ran a heat up test. I started the car and let it idle for 20 minutes. The back end of the cat started at 70 F and heated up to 450 F. The end of each wing where they attach to the original bracket supporting the ends of the muffler both moved to the back 0.25". The distance between the outboard ends of the two wings also increased by 0.1". I still have trouble understanding the magnitude of the growths I measured but the data is the data.

Certainly I need to test this design thru many heat cycles. I like the heat growth movement that it supports but I do not like the stiffness in the vertical direction. I am going to try a different design that would allow for vertical motion but keeping the tips centered is an ongoing concern.

From a cost and effort viewpoint this design is pretty cheap. I cut out the stainless on a band saw and filed it smooth. The holes were drilled with a hand drill and slots elongated with a file. I bent the stainless with a brake but could have done it in a vice. I also hand cut and drilled reinforcing plates for the wings where they bolt to the center support.

I welcome any and all comments.

 

Just realized my pictures were taken before I put the bolts in at the ends of the wings. Yes, I did put the original bolts in those holes.

 

For anyone that followed my saga of trying to build a better support for the exhaust system, here is an update. Not long after I built my modified support system I decided I was putting patches on a poor design. Looking at exhaust systems in a variety of other cars, the support system should support the weight of the exhaust system but allow it to move relatively freely due to thermal expansion. With a single muffler joining the cats from either side of the car, the system is over constrained and it somewhat limits how the pipes can grow as they heat up. Removing all but the rear most support bracket and softening it up in the fore-aft directions helps but is limited in how it can help. To me, the best solution is the sport exhaust which treats each side of the exhaust system completely separately. So I switched to the sport exhaust. Each side can now move independently. The system is supported at the back by soft mounts so expansion is pretty much unconstrained. As a data point, after a spirited drive I measured the fore-aft position at about where a capristo bracket would be. I then let the car cool off for many hours and remeasured the same points. Going from hot to cold, the left side moved forward about 0.2 inches and the right side moved forward 0.25 inches. Note that these numbers are about 0.05 inches less than I measured with the stock exhaust. Unfortunately I did not come up with a great solution to the stock exhaust. However, I am very happy with the sport exhaust setup.