[BigD]

Hey guys,

I'm working on a new turbo manifold for my car and I'm doing the runners out of 321. Experienced turbo racers told me to use a mild head flange for expansion and to weld to it with 309. Burns seems to agree but I was wondering what you folks thought? I know 309 is meant to weld 300 to mild to compensate for chromium dilution but the reason I'm using 321 with 347 filler is because I will regularly see over 1500f, the temperature at which 308 welds can suffer from carbide precipitation. So if I'm using 347 to stabilize the weld, why not use it for the head flange too? The Lincoln book seems to suggest that you can use 309-Cb for a root and then 347 for a cover pass. I don't have 309-Cb but maybe I can do a 309L root and 347 cover?

Thanks guys!
D

[kiwi2wheels]

No temperatures are mentioned in post #4 , but there is a PDF link to the paper.

https://forum.millerwelds.com/forum/wel ... -stainless

" 07-20-2011, 05:33 PM
We use 625 at my work welding carbon to stainless boiler tube. They stopped using 309 years ago and haven't had any crack related issues since going to 625. Back when they still used 309 in this application they had alot of issues and came up with the EPRI prep. EPRI stands for electric power research institute . It is a collaboration of electric power co-ops and manufacturers of this type of equipment to come up with better and more cost effective solutions to power plant issues.

Development of a New Nickel Filler for Dissimilar Metal Welds and Repair
Category: Technical Results Published: 7/9/2009 Product ID: 1018991
Type/Size: Adobe PDF (.pdf), 19.20 MB Program: Fossil Materials and Repair

Abstract: In the late 1980s, the domestic utility industry suffered from dissimilar metal weld (DMW) failures between low-alloy ferritic tubing and austenitic tubing in superheaters and reheaters. The Electric Power Research Institute (EPRI) performed extensive research into the problem and found that nickel-based filler metals showed significant service life improvements over 309 SS filler metals. Furthermore, improved joint geometries and additional weld-metal reinforcements have provided added service life. To ... "

I'm curious, what reason(s) were you given to use mild steel flanges ?

[BigD]

Never heard of that and I don't doubt it works but I'm specifically after extreme temperature resiliency (hence using 321+347). Even Burns who makes pro manifolds says to use 309 on the header flange, and I'm sure it's fine as that joint isn't as critical as the tubing butt joints or the collector but I still think I can do better with the addition of 347 but as usual, good intentions without education/experience can lead to bad decisions.

Mild for the head flange due to thermal expansion rate (the lower the better when you're talking about the thing binding the runners together).

[Poland308]

We’re talking exhaust with a thickness of .1???. Use 309. You would only use a high Chrome filler over top if you were getting into thickness over .5.

[BigD]

Schedule 10 1.5 yeah so .1" wall. It's not so much about the amount of chrome but the Cb of 347 to stabilize the joint and prevent carbide precip (in my mind anyway).

[Poland308]

One or two passes of 309 is more than enough. And it’s good practice.

[Arno]

Flange/tube interface should also be one of the less hot sections of the manifold as the cylinder head (with it's cooling passages) forms a massive heatsink on the flange itself, so I'd expect 309 to do fine.

Just make sure the ports on the head match (or are smaller) than the ports/openings on the manifold flange and tube diameter. That way your'e not ramming hot gasses right into the wall of the tube straight from the exhaust port (not to mention it setting off a terrible swirl in the manifold that wrecks flow)

Hottest area of a manifold tends to a be a little away from the exhaust ports and near the turbo where the gasses hit the impeller and the momentum gets lower so the heat transfer from the gas to the wall(s) happens more.

Even on N/A engines with tuned length headers you'll see the glow start a little downstream and at the merge sections. The flange connections stay pretty dark.

Eg:

https://www.youtube.com/watch?v=RgPHKwoo8fk

https://www.youtube.com/watch?v=YPtD_oiFqPg

If you're expecting a lot of heat in the manifold and using 321 then I do hope you have a water-cooled (external) wastegate on your setup or you'll be using them as consumables

Bye, Arno.

[BigD]

Ahhhhh yes, that's the missing link in my thinking - the heatsink. Thank you Arno, makes perfect sense now.

My wastegate is on a takeoff from the turbine housing, fairly remote, and it's a big ole V60. They're not watercooled but it's in a good spot in terms of cooling