IW STi Forum banner

1 - 11 of 11 Posts

·
Registered
Joined
·
644 Posts
Discussion Starter #1
I have a rotated twin scroll exhaust manifold with external wastegates so it has a lot of flanges and the manifold is 8 years old so the flanges are warped. I spent the last month sanding every flange flat and finally it is good as new! I have a smoke machine that I use to find exhaust leaks and when hooked up to my turbo manifold it now shows no leaks including the v-band connections for the wastegates. I then hooked the manifold up to my air compressor to see how much psi it can hold, when I put air into it it would only build 2psi before most of the flanges and v-bands would leak.

Which had me wondering if that is normal? What exactly is the concept behind exhaust manifolds and their ability to hold pressure. Are the flanges including the v-band ones torqued to proper spec with proper gaskets designed to be air tight when pressurized? If the pros encounter a warped exhaust flange causing a pre turbo exhaust leak once it is fixed when do they stop, do they do a smoke machine test and if no smoke comes out is it considered good, or do they try and build an amount of psi on the flange to see how well it is air tight.

Keep in mind I am doing all of this with the exhaust manifold off the car assembled in my garage with each end capped off.
 

·
Registered
Joined
·
118 Posts
Keep in mind I am doing all of this with the exhaust manifold off the car assembled in my garage with each end capped off.
How tightly are the ends capped off? There may be a difference in clamp force between actually being installed in a car vs a bench setup. Do you know how much pressure was being pushed through with the smoke machine?

Also, consider a that there will be certain amount of thermal expansion in all the flange and v-band points when it's running on the car, thus more tightly sealing the connections. There would also be gaskets to help seal the flange parts.

I can't give you an exact figure for exhaust manifold pressure, but it's definitely higher than two psi.

Sent from my SM-G955W using Tapatalk
 

·
Registered
Joined
·
5,449 Posts
V-bands should not leak, ever. They are however, very prone to warping if the process used to weld them is not on point. Take a flat edge and go across the flange in several directions and you can see if it's got a wave in it, or is cupped. We test our headers to 25psi and if there's even a single tiny bubble, the v-bands get cut off and rewelded; which is the only way to properly fix the issue.

Some things to check are making sure hte v-bands are seating concentrically to each other via the centering ring/boss. That you are tightening to the proper torque spec in in/lbs, and that there is NO debris (not nick or scratch) on the flange faces.

In use, you will typically see 2-2.5X boost pressure in the manifold. TS is less, except for high RPMs where they are more restrictive than thier SS counterparts.
 

·
Premium Member
Joined
·
5,389 Posts
V-bands should not leak, ever. They are however, very prone to warping . . . We test our headers to 25psi and if there's even a single tiny bubble, the v-bands get cut off and rewelded; which is the only way to properly fix the issue.
Chris - OP took material off - did that make them useless . . . seems your answer is yes since "if there's even a single tiny bubble, the v-bands get cut off and rewelded; which is the only way to properly fix the issue"
 

·
Registered
Joined
·
5,449 Posts
Chris - OP took material off - did that make them useless . . . seems your answer is yes since "if there's even a single tiny bubble, the v-bands get cut off and rewelded; which is the only way to properly fix the issue"
I had assumed that material was removed from the other flanges; manifold-to-head. If the v-bands were sanded, then they need to be replaced.
 

·
Registered
Joined
·
644 Posts
Discussion Starter #6
My smoke machine is a smaller one with a 1 psi restriction, so it only goes to that. During a bench test I am torquing my block off plates to 28 ft-lbs; I haven't looked up the torque spec for a stock manifold but I know 28 ft-lbs is close to that. Killerb I'm glad you chimed in I was hoping you would give your opinion, I did sand my v-band flanges and was aware that it may ruin them however they were warped to begin with and they hold air a lot better sanded then previously warped so I was willing to take the risk. To be clear too I am specifically talking about 38mm ewg v-band flanges with the fire rings, I know there are different styles of v-band flanges that our cars use so I don't want to have a mix up with the style of flange. Having the flange sanded flat seems to of seated the fire ring much better, as if the warp-age was preventing the ring from fully seating.

Prior to making my own block off plates to bench test my manifold I was trying to find a proper torque spec for a tial 38 mvs wastegate and I could not find anything, so with the smoke machine hooked up I tightened the v-band clamps and at 4.5-5 ft-lbs they stopped leaking. If all else fails I plan on sticking to that torque spec as I do not want to have them too tight is there a suggested torque spec for a ewg v-band clamp?
 

·
Premium Member
Joined
·
5,389 Posts
Regardless it didn't hurt to try - they weren't sealing warped and non-flat . . .
 

·
Premium Member
Joined
·
5,389 Posts
there will be certain amount of thermal expansion in all the flange and v-band points when it's running on the car, thus more tightly sealing the connections. There would also be gaskets to help seal the flange parts.
Don't the clamps expand by the same amount negating most increase in in clamp pressure - there may be a small increase, perhaps tiny, because the clamp is probably a bit cooler the the flange, but not a great percentage of the total rise.
 

·
Registered
Joined
·
118 Posts
Don't the clamps expand by the same amount negating most increase in in clamp pressure - there may be a small increase, perhaps tiny, because the clamp is probably a bit cooler the the flange, but not a great percentage of the total rise.
Depends on materials used. Not all metal expands at the same rate at the same temperature. Even if the clamp and band section was similar material, you still have the clamp force putting pressure on the joint. That being said, it depends how big the leak space is. While the metals will expand and the space will reduce, it may not be enough to fully seal.

Sounds like OP figured it out though. Just torqued down the clamp a bit more. If that's all it took, the leak couldn't have been that big.

Sent from my SM-G955W using Tapatalk
 

·
Registered
Joined
·
644 Posts
Discussion Starter #10
I am on the side that metals expand and contract at different rates based on the material and circumstances. I reached out to Tial and asked if the V-Band clamp on a 38mm MVS wastegate has a torque spec and it does they said the clamp should be torqued to 7-9 ft-lbs which is some useful info. I torqued mine to 4.5 - 5 ft-lbs so I shall do another bench test with a proper torque. Not only that but I plan on making a little gasket if the seal is not complete to see how that does.
 

·
Premium Member
Joined
·
771 Posts
I think sealing or testing the exhaust under pressure is misunderstood. Post turbo exhaust pressure is measured in a few PSI. 2-8 over the RPM range on a relatively open exhaust. You don't want exhaust leaks. They can cause backfiring and might even asphyxiate you. But, the post turbo pressures are low.
Pre-turbo pressures can be quite higher, usually they are 2:1 from what I've read.
Ex.: 40 PSI @ 20 PSI boost.
 
1 - 11 of 11 Posts
Top