Subject: exhaust flange-pipe interface failures
From: ACRE NL
Date: 1/10/2003, 7:02 PM


Here is another way to keep stress off of the flanges. A slip joint just
 > > outboard of the header flange. The tabs should have heavy springs
pulling
 > the
 > > two halves together. For aircraft use, an additional hole in each tab
 > inboard
 > > of the springs will allow for a 4130 strap to connect the inboard
holes.
 > > Holes in the ends of each strap align with these holes in the tabs. AC
 > bolts
 > > with jet nuts go through the straps and tabs and are just torqued to
just
 > > short of tight. The pipes can move around during operation.
 >
 > > This is used on race cars around the world. Another method is to have
the
 > > tabs long enough to pass each other, and a hole through both tabs lines
 > them
 > > up, and a single bolt does the same thing. My muffler weighs 10
pounds.
I
 > > wouldn't dare hang it on the engine. Failure of a tube would be
immediate.
 > I
 > > did that already. I also hammered a SuperFlow muffler to pieces. I also
 > > melted the end off of a steel bodied steelpack muffler. I also melted a
 > piece
 > > of tunnel sheet aluminum and overheated a Halon fire bottle. It went
off
 > > during the race and scared the bejesus out of the driver. I also melted
 > the
 > > outside of a tube bend and boiled the fuel out of the carb, even with a
 > > shield in between. The header tubes squirm around during operation.
They
 > > don't move far but have a great amount of force. The tubes have to be
free
 > to
 > > move. The muffler weight and even the header weight cannot be hung on
the
 > > engine. They are running bright orange at 50 degrees lean of peak EGT.
 > Even
 > > stainless looses some strength at 1,700 degrees.
 >
 > Lynn
 >
 > Lynn;
 >
 > What alloy do you guys use?  SS (like 321) will fail much more easily at
 > 1700 than Inconel; especially if under any stress loads.  It gets
 > inter-granular stress corrosion and loses it strength.
 >
 > Al


I have a few sets of CART style headers from Rahal's car. They are Inconel
but too large at about 2 1/8" I would love to use that stuff. But they are
changed out after three heat cycles because they crack. They are very thin.
After melting a .120" wall tube bend right outside the port, I went to .135"
wall bends close to the engine. This is carbon steel ERW with the seam on the
inside of the bend. Also the outside of the bend gets pulled down to a very
thin wall. So, .135" wall through the first bend, and then .120" wall to the
collector. The collector is .062" mild steel, then a slip joint and a 14"
megaphone up to 4" slip joint at the 4" Borla stainless muffler. So starting
with say, .062" stainless and then bending it makes it very thin. Another
thing that can take away some stress, is keeping the run from the port
straight out of the port for as great a distance as possible before the first
bend. This also improves power, and is a nice place to install the EGT
probes. The engine builder (Daryl Drummond) likes them three inches out from
the port. Better in a straight run than in a bend where the outside of the
bend is getting most of the flow.

   In stainless, I would build it the same but in .062". Never put heat tape
on any system close to the engine. Get some air on the exposed headers close
to the port. Then wrap the stainless as much as you want. No heat tape at all
on carbon steel headers. When the Pport header started to leak after a few
weekends (heat taped all the way up to the flange) the heat tape had a little
hole in it. So I thought I would untape it and weld the hole up at the track.
Most of the pipe was gone. The tape was holding hundreds of little pieces of
steel together. So we raced home and built another set of headers and made
the race the next day. Without heat tape.

    The CART pipes are made from many small pieces if Inconel sheet stock
that is formed to produce the bends and hand welded together. The outside of
bends are gathered a bit instead of being stretched and made thinner.

    Burns would know the right alloy to send you. TIG weld all of your
stainless. Weld only the engine side of the flanges. Four spot wells on the
header side seems to be the way to go. I would bolt the flanges down tight
with GE silicone tub calk for a gasket. The a run of tube in line with the
port and a slip joint. Then equal length tubes to the collector with one or
two slip joints at the collector. Then the stainless muffler mounted soft.
Then the tailpipe overboard, mounted soft. Belt sand the flange faces then
lap them on a piece of glass and a piece of oiled wet or dry paper.

Lynn E. Hanover

Good advice Lynn. Here is a bit more info while we are on the subject.
Ed Anderson is using .140 thick SS pipe bends designed for cheap SS hand rails
and it does'nt even discolor. A bit heavy. You can buy it from RB Wagner.
Ideal to build your own turbo manifolds. I would only use it for the
first couple of feet and then go down on the wall to .065 and finally
.035 out on the cooler ends. Slips joints on all transitions.
A system like this should last forever.

Paul Lamar
 
The AirCraft Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/            
Copyright 1998-2002 All world wide rights reserved.