Where to get butt weld fittings for DIY manifold?

[Can'tWeld]

Hi,

I am planning to build a log manifold for my turbo MX-5 project.

I already have a flange, so now I just want to buy some fittings to make the manifold itself. Ideally I'd like two 1.5" long radius 90 degree elbows and two 1.5" equal length tees to match a design I've seen on a forum.

I've found one supplier of these in mild steel but it seems really expensive (comes to about £75 and that's before postage!)

I've found them cheaper in Stainless but my main worry there is that apparently I'll struggle to crush the ends of the tube in a vice to match the ports on the header flange, is that true? Also I've never welded stainless before but that's less of a problem.

If stainless is a no-go, can anyone recommend a good supplier of mild steel pipe fittings who can sell me these bits without it costing an arm and a leg?

Cheers!

Rowan

 

[normspanners]

Tee's on an exhaust manifold ?

I've found them cheaper in Stainless but my main worry there is that apparently I'll struggle to crush the ends of the tube in a vice

?? how thick section are you using ?

 

[Can'tWeld]

This is the design I was planning to try and replicate http://www.mx5nutz.com/forum/index.php?showtopic=89841

I've been told elsewhere tees are a bad idea - though I can't see the flow of that being that much worse than doing it any other way, they do seem to be a pain to find, so perhaps I'll make it out of elbows only

I was planning on using either sch. 10 or 40 tubing, so I shouldn't have any problem forming the ends of this in a vice if it's SS?

Thanks!

 

[WorkshopChris]

Looks like that should be a replacement part for an old generator or something sitting in a yard chugging away year after year, not anything performance.

 

[normspanners]

they are steam / hydraulic bends !! no wonder you would have difficulty squishing them
speak with http://www.protechsupplies.co.uk/contact-us/ they are just around the corner from me and have it all in stock but I would not use the heavy gauge stuff.
They have 1.5 or 2mm section stainless in stock also, and they are classed as dairy fittings and loads of people use them for exhaust bits, I know I do !

 

[normspanners]

You will increase the pumping effort using tees on the exhaust stroke but a turbo does that anyway by increasing back pressure before the impeller vane. So I suppose it is a lose some gain a lot situation, but I'm sure it could be done alittle more flowed than that, as that setup does look a little like a diesel industrial engine design that has max air flow into each cylinder EVERY stroke,
The thing is with each 'leg' being so short, some of the pressure will tryand push dead exhaust gases back into the next opening exhaust port ruining the natural flow, and therefore the whole idea, of utilizing the full turbo effect by INCREASING the exhaust flow would benefit from a more flowed manifold IMO
A little like this where each leg allows enough space for each exhaust pulse and the venturi effect 'pulls' on the next opening valve

 

[CompoSimmonite]

Looks like that should be a replacement part for an old generator or something sitting in a yard chugging away year after year, not anything performance.

Reminds me of the type of manifolds you see in car museums on pre war side valve engines

 

[mr migwire]

im having a bit of fun off and on making a 2 into 1 adapter for an mr2, we picked up a nice balanced 4 branch for it and i stupidly agreed to knock up the adapter , all good fun.

 

[Pete.]

Norm's right, tees on a turbo manifold is a terrible idea. Whilst equal-length runners are nowhere near as important on a turbo manifold as on a tuned NA exhaust, you don't want to be firing exhaust pulses either across the face of the other runners or back up towards other exhaust ports. the log-style manifolds you see often do this and it's very poor practice.

 

[mechman]

Are swept tees not available for your application .

 

[normspanners]

Just some interesting reading ;

http://www.s2ki.com/s2000/topic/1000188-my-spool-results-of-log-vs-tubular-manifolds