Designing and manufacturing of exhaust pipes

[Sketchy_Racer]

After reading some of the content in the ESE thread, I've become really inspired to take a closer look at hydroforming chambers. In particular for our NSR300s which we are currently using the Tyga chambers but are having lots of reliability issues with them and also I am sure that there are gains to be made.

So I had a good look at the link to the Pitlane thread that Husa put in ESE See Here For Link to Pitlane

There are some gems of information there as well as a few photos which I've attached here. From what I have read so far there is a large importance placed on the smoothness of any of the transitions in the pipe. And sharp changes in angles will cause wrinkles when pumping it up. Second important detail is to roll the edge of the two halves of the pipe so that you can butt weld them. If you don't do this it will cause the pipe to try tear the weld apart when pumping it up. It seems that the general consensus is that gas welding the two halves gives the best results, probably something to do with the spread heat of gas welding that will help anneal the metal as you weld as opposed to tig welding which has a very intense but small area of heat which can cause the creation of a brittle martensitic structure in the steel if there is enough carbon content.

I have my sheet of steel turning up on Monday so will get straight into it. I am going to start with my first pipe using Frits Overmars guideline for the pipe design. I am very limited on time for this project (work and study is a killer!) so any input on the actual design would be much appreciated

For getting the flat pattern I am going to give a new feature in solidworks a try. In the 2015 version I have the ability to model the whole chamber as a surface then flatten that surface. This tool was developed for the textile industry where users need to flatten complex curves into 2d paters (think shoe manufactures getting leather cut as flat) So I am hoping that the same principle will apply for the flattening of the chambers.

[Bert]

Nice

We have been playing around with this concept again. Scottie use to have some amazing hydro formed pipes back when we first started buckets last century... He has built a cool tool to make the physical process a lot easier.
Anyway I was directed to
Ezin Hobeki - pipe designer software
Outside of doing some magic two stroke calcs - you can load in your own specs. It has a good hydroforming construction tool allowing you to generate what ever you like (DXF file output).

cone designer

A9 cad consuming the DXF output.

After contacting Ezin he added in a little more flexibility so I could design four stroke megaphone (using the pipedesigner software and outputting to DXF files to tweak in A9Cad freeware).
Either send the DXF file to the water cutter or print out on paper and cut away....

[Drew]

Subscribing.

[190mech]

Glad to see this thread!I did a hydroformed tapered headpipe for a quite pipe design that needed extra length for lower RPM torque.I found that the bends form tighter than the pattern and must be adjusted because of this;

[Crasherfromwayback]

Love yer work.

[Grumph]

Ideally, you'd use deep drawing steel sheet. After much asking, it seems that it's no longer imported into NZ - at least for public consumption.
I finished up using bright mild 20G and annealing all over once it was welded.
I was looking for a way to mass produce them so investigated continuous arc welding at a company here in ChCh. Afrer truing up their wheels - and adding a fence to one wheel to keep the weld seam aligned with the edge...Fail. Even with the amps turned up to max the seams split.
So back to TIG...Which worked.

I used a hand pump. Marsheng on here imported a box of them. Used with a one way valve in line they work very well indeed. I'm pretty sure he's still got a couple of them if anyone is interested.
I also welded in fittings at each end. Steel pipe at the input end to suit the one way valve. A machined stub at the exit end tapped for a small lever tap.
Pump through till no air, close the tap and pump...Go back and reweld splits, resume as before....

[AllanB]

So you weld up the 'flat' unit then pressure fill it with water to get the roundness desired?

[Drew]

So you weld up the 'flat' unit then pressure fill it with water to get the roundness desired?

Seems easier than what I imagined hydroforming to be.

Now it makes sense where someone said to Chop the seam off after forming, and butt welding afterward.

[Sketchy_Racer]

After a little bit more research I have found some more notes and photos.

From the "Practical Machinist" forum

I visited Pete Gibson when he was in Romford, UK in 1994.

I flew all the way from South Africa so that I could learn how he hydroformed an exhaust. I will outline the common mistakes. I used this technique for about two years making racing 2-stroke exhausts.

Common misconception is that the process is dangerous. Water can't be compressed like a gas can, so all that will happen is that the water will fall out and make a mess.

The trick is working out what radius the exhaust will end up once it is formed. Trial and error will give you an idea what to allow for the radius tightening up.

You also need to allow about 2% extra for stretch during forming.

If you TIG weld, use thin mig wire to prevent porosity in the mild steel. Gas welding is preferable as the weld is not so hard after welding.

The edge of the two halves needs to be bent over using a bead roller or sheet metal Jenny. This gives you the correct joint for welding and what it will end up after forming, butt weld and not an edge weld if you welded them flat together.

Allow extra both ends for attaching the water pump and a bleed valve.

I used a old hydraulic jack which I cut in half and welded fitting on to the various inlet/outlet holes of the jack. I used an old motor cycle brake line as the hydraulic line. Only issue with this choice is it takes ages to fill the exhaust before it starts to form.

Fill the exhaust with water, and bleed all the air out using the bleed valve. Now you need to stop the large area expanding first by placing the exhaust in a hydraulic press between two steel plates.

Gradually expand the exhaust, making sure the small areas obtain correct shape before you fully expand the large section.

You will reach a point where the exhaust has taken final shape but has unacceptable wrinkles. This can be dressed out whilst under pressure with either a plastic THOR hammer or steel planishing hammer.

Heat can be applied to the wrinkles to make them expand out. You will need to drain the exhaust to apply heat, otherwise the water will soak the heat out. Once you have applied heat to the low spots, re pressurize the exhaust and the heated areas will expand first.

I also experimented with making steel dies for the large parallel section to give it better definition. Make two halves in the lathe and make them so that you can bolt the two halves together. This will give the exhaust better definition in the areas where the exhaust would normally end up with a radius.

[AllanB]

Very interesting method. Done well the final result looks impressive. It appears to be similar in 'work' to welding up a pile of cones. Any measurable difference in the final product on the bike?

[husaberg]

I stole this off the Pitlane thread where your photos came from
Howard should be contactable via a few sites I might ask him to post something more
The major points I gleaned were
The bending of the edges prior to joining
The amount of stretch
The banding just areas can only grow so much
The way he coaxes the ripples out.

I beg to differ! I have developed a technique (sort of a reverse engineering) that allows me to hold a tolerence of less than .5mm anywhere on the pipe. My hydroforming technique produces a ripple free pipe that does not have any sharp edges that can disrupt the wave. Even if a pipe is lazer cut and uses computer generated layout there is still an abrupt change in direction everytime the wave transitions from a straight section to a curved section.
I use simple tools a torch a pressure washer and a hammer. The metal I use is 19 gage .041 forming metal. The secret is in the design of template and how it is blown up. On the bottom of page 40 of this thread there is a picture of a pipe I built using this process.

Hydroforming is not easy but with good forethought and good craftsmanship along with good process results can be repeated with good consistancy. I use an accurate weigh scale to know exactly how much water is in the pipe so I can duplicate the volume of each pipe exactly. I use a lot of construction lines in the prototype stage and measure each line circumference to see how far off the drawing the pipe has stretched to. I then adjust the template accordingly until I get the results that match the drawing within .5mm circumference. I can also vary the volume with the same design to test different volumes. I record the weight so I know the exact volume of each pipe and test each design to find the sweetspot. It is both a science and an art

My investment in making a pipe is extremely low compared to a CAD designed stamp. A sheet of white cardboard is under a dollar and is what I invest in the template. My time of course is an investment but I would rather design a pipe than sit in a pub drinking beer complaining. I was taught how to hydroform by a very good friend by the name of Alex Mayes. He was a craftsman and reminds me a lot of Frits. Unfortunately he had a bad heart and he was a chain smoker. When he suddenly died I was left to either find another pipe builder or start making my own. By the way frits your first try at hydroforming looks surprisingly a lot like my first attempt. I have got very wet a few times and it took about a year before I could say my pipes were as good as Alexes pipes. I have probably made 200 sets of hydroformed pipes since and every time I finish a pipe I get excited to try it. I will gladly answer any questions if anyone wants to attempt it.

The difuser usually starts out as a 3 stage straight cone design because that is what my software spits out and that is how nmost pipe designs begin life. I make a full size straight line drawing of the actual finished pipe. I then use a french curve to blend the intersecting cones so as to make a smooth transition. Sometimes I blend the cones almost 1/3 of their entire lengths to make a more or less parabolic curve. I have found that a straight cone difuser at near the center section delivers the best powerband. A more parabolic continuous curve gives a more peaky power band with a little more peak horsepower. Just to let you know the template drawing is typically 1.5" smaller than the finished diameter. The metal stretches that much. That is the secret to getting very smooth finished products. Also to achieve high pressure in the pipe I use sleeves that exactly match the finish diameter of the center section. This only works on pipes that have a defined straight belly section. THe pipe expands into the sleeve allowing me to build pressure in the rest of the pipe to remove any ripples or dents. For a pressure source I use a standard piston type pressure washer with a standard wash handle fitted with a hydraulic fitting. It is vital that the handle be leak free and is able to control the flow. it is also vital that there be no pinholes. A lot of the time is spent with a pressurized pipe on your lap hammering away with a small ball peen to dolly out the ripples. With about 300 - 400 psi in the pipe it allows you to work dents and ripples by tapping around the dent. It is also vital (and may save your life) that you ensure there is no air at all in the pipe when blowing it up. Even a small trapped pocket of air can store enough energy to do serious damage if it were to rupture. With no air the rupture gets you wet, thats it! People ask me about the weld. If you can't weld then dont try to blow up a pipe. Practise and get good at welding with acetylene first. I have tried several types of filler rods and the best I have found by far is 19 gage "Mechanics" wire. Yes the stuff used to twist things together. it is cheap and most importantly it is malleable even when welded. The outside seam actually shrinks quite a lot so you need a soft weld that will shrink with the metal
I use 19 gage (.041") cold rolled steel ,nothing special.
I use a home made hand driven roller to roll the edge. I also use a small press that has an adjustable relief valve to allow the pipe to expand automatically in the press. The prssure setting depends on the size of the pipe . I press the pipe between two plates of steel covered with plywood. The press allows the entire pipe to expand evenly. Without it the center expands first and creates some horrible ripples. Everythin g is done cold. I do not recommend heating the pipe at all with any pressure in it.

The templates are made "old school" on a drafting board. I make a centerline using an aluminum yardstick bent roughly the shape of the bend of the pipe and mark it off in 1 inch segments. I transfer the measurements from a full scale straightline drawing. I connect the dots with a flexible curve. I number each segment on both the template and the drawing and transfere the construction lines to the metal for future reference

[Lightbulb]

Ideally, you'd use deep drawing steel sheet. After much asking, it seems that it's no longer imported into NZ - at least for public consumption.
I finished up using bright mild 20G and annealing all over once it was welded.
I was looking for a way to mass produce them so investigated continuous arc welding at a company here in ChCh. Afrer truing up their wheels - and adding a fence to one wheel to keep the weld seam aligned with the edge...Fail. Even with the amps turned up to max the seams split.
So back to TIG...Which worked.

I used a hand pump. Marsheng on here imported a box of them. Used with a one way valve in line they work very well indeed. I'm pretty sure he's still got a couple of them if anyone is interested.
I also welded in fittings at each end. Steel pipe at the input end to suit the one way valve. A machined stub at the exit end tapped for a small lever tap.
Pump through till no air, close the tap and pump...Go back and reweld splits, resume as before....

Is Greens industries who used to make wheel burrows still in business. They would have been made from such steel. I will have a phone around on monday. Can this process also be done with Aluminium ? if so, what grade could anybody suggest please. I find this all very interesting.
Neil

[Grumph]

I note he says early on, "19 gauge,.041in forming metal"
Then later, "19 gauge cold rolled steel, nothing special"

In my limited experience, you don't see a lot of stretch, if any, with normal cold rolled mild steel (20G, or thinner than his...). You might very well see it with 'forming metal' which i take to be deep drawing steel.

If he's still around Husa, and you can contact him, it's worth clearing up just what steel he is using.

[Grumph]

Is Greens industries who used to make wheel burrows still in business. They would have been made from such steel. I will have a phone around on monday. Can this process also be done with Aluminium ? if so, what grade could anybody suggest please. I find this all very interesting.
Neil

You'd need to find someone doing a lot of presswork. Do Fisher and Paykel still do any in this country ?
I'd doubt if alloy could be used, I'd think it would tear too easily in the light gauges used.

[jasonu]

You'd need to find someone doing a lot of presswork. Do Fisher and Paykel still do any in this country ?
I'd doubt if alloy could be used, I'd think it would tear too easily in the light gauges used.

I'd be surprised if F&P did anything more than assembly in NZ.