Manufacturing of Plenum

[Gaanja]

We are having problem manufacturing the plenum. Its an organic shape with curves and fillets.

Last year we tried aluminum sand casting, but the weight and the thickness was too much.

This time we tried to get it done by bending sheet metal. We used 3 mm thick sheets so that the plenum is stiff enough and would not breathe when the engine is running. But the contours had tight radii and we could not get the right shape.

i would like to know if there is any other way we can get the plenum manufactured other than carbon fiber or RPD. We don't have the resources for that.

Has anybody tried a plastic injection mould??

[Gaanja]

We are having problem manufacturing the plenum. Its an organic shape with curves and fillets.

Last year we tried aluminum sand casting, but the weight and the thickness was too much.

This time we tried to get it done by bending sheet metal. We used 3 mm thick sheets so that the plenum is stiff enough and would not breathe when the engine is running. But the contours had tight radii and we could not get the right shape.

i would like to know if there is any other way we can get the plenum manufactured other than carbon fiber or RPD. We don't have the resources for that.

Has anybody tried a plastic injection mould??

[Auerbach]

I've built plenums out of a combination of sheet and tube but nothing near 3mm thick. I think we used .046 Al or something. Unless you have large flat surfaces it shouldn't breath enough to affect performaces. If it does, maybe you can add a stiffening rib on the outside? Not pretty but it could be effective.

[Simon Dingle]

Sheet metal work really isn't suitable for complex surfaces incorporating tight radii.

I agree with auerbach that 3mm is too thick, last year our plenum was made from 2mm thick low grade aluminium.

Finally, does your intake manifold need to be a complex shape? How much power are you aiming to produce? If it's less than 90bhp, then I suggest that you don't need to have complex curved surfaces and that it could be made from multiple flat surfaces.

[exFSAE]

DFM. Learn it, love it.

If you don't have the manufacturing capability to make crazy shapes with tight radii.. don't design it like that.

[Mikey Antonakakis]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by exFSAE:
DFM. Learn it, love it.

If you don't have the manufacturing capability to make crazy shapes with tight radii.. don't design it like that. </div></BLOCKQUOTE>
I agree heartily.

Here is how I made this year's plenum:
-This plenum has complex bends in the runners, three bends per runner, and they are pretty short.
-The plenum is shaped like a vase. The runners are interfaced with the plenum in a square pattern.
-First, I milled out female molds for the intake runners, in two sections because I have out-of-plane bends in the runners.
-I assemble the molds together and cast wax into them to create a male mold.
-Then I fiberglass over this. When it cures I melt the wax out.
-Next, I rapid protoyped the bottom surface of the plenum. However, I could have gotten away with machining this... but because I had access to the machine, I made a more complex geometry.
-I then reinforced that piece with fiberglass.
-Next, I rapid prototyped the plenum. Again, I would not have done this had I not had access to the machine. I would have turned down a large cylinder of wax, fiberglassed over it, and either melted out or removed the wax depending on the shape. In my design the wax would not have been captive.
-Anyway, then I reinforced the plenum with more fiberglass.
-Next, I fiberglassed all of the pieces together.
-Then I found a set of stock throttle bodies, and removed all material except the injector and fuel rail mounts using a manual mill.
-Then I bonded that to the manifold.
-Finally, I made a flange for our throttle body and bonded it to the plenum.
-Checked for leaks, assembled onto the engine, and ran for the first time just a half hour ago.

Now, this process took me at least two weeks with help from a partner with the fiberglassing, and working several hours every day. It would have taken a lot longer had I not had access to the rapid prototyper. Personally, this is the absolute limit of complexity I'm willing to deal with in the construction of the manifold, but I'm very happy with how it turned out and I don't feel that I had to make many significant sacrifices in my design for the sake of manufacturability. But again, this is my second intake manifold, and I have been working with fiberglass for a while now, and I'm very stubborn and try to hold myself to high standards when I'm in charge of something. I wouldn't recommend this method unless you are equally as crazy. I would use a simple log or triangular plenum. Some very successful teams use fairly simple plenum geometry. I personally like making it with fiberglass because of the light weight and because I can't weld aluminum very well. Wax is easy to work with and doesn't stick to epoxy.

But, as exFSAE said, DFM. I did it to the absolute limit of what I can personally tolerate in terms of manufacturing headaches, and it was worth it for me. I don't think you'll need incredibly complex shapes. My '09 plenum is maybe 1.5mm average thickness of fiberglass, and only broke once, at a huge stress concentration, when I sprayed way too much starter fluid and left the throttle closed. So I think 3mm of aluminum is probably too much... how much would that intake weigh?

[Gaanja]

Thanks for all those opinions.

This is the shape of the plenum

<a href=http://img17.imageshack.us/img17/8310/73107579.jpg"> </a> Uploaded with ImageShack.us (http://imageshack.us)

the 08 car had a plenum which was box shaped curved on the vertical axis by 40 degrees. it was done by rolling and welding sheet metal. It was 2 mm thick and we could visibly see a change in contour after testing. that is why we wanted to use 3 mm plates this time round.

We also thought of hydro forming but that will turn out too expensive to make just a few pieces.

[shivam popli]

Sir,
May i know that what is the material used in RAPID PROTOTYPING of plenum and if that material has the required strengh.I want to now one more thing that materials like silica glass would be durable for making intake runners or not.

[Wesley]

That's very similar to the shape of the '08 intake manifold I designed, except ours dumps into a cylindrical plenum.

I ran into similar issues once I realized we couldn't make it out of carbon (money and time,) so I ran a few calcs and realized that just squaring it off instead of filleting those radii made it significantly easier to manufacture and didn't impact the power by much at all, in fact, we made our highest amount of official power that year (72HP. Not that much but it was a fat curve http://fsae.com/groupee_common/emoticons/icon_wink.gif ) and that was simply with intake and exhaust, no engine modifications.

Here's what I suggest. get rid of the fillets, and make it with 90º joins. To visualize this, look at it from the top projection, and just extrude that projection as a solid. That's the shape you want to make.

Weld it up and then get a piece of round stock and a hammer and form the squared restrictor opening to the proper size. You may have to stretch the metal a little bit. We made ours out of .035 aluminum without issues.

The benefits of the fillets are pretty marginal. If you're worried about the look, you can just grind your welds smooth (just be sure they penetrate!)

[carlosm]

i have to say making an intake from carbon isn't that hard or expensive, we managed to crank ours out of carbon and aluminum.

http://arizonafsae.com/Univers...y/Pages/2010.html#30 (http://arizonafsae.com/University_of_Arizona_FSAE/Gallery/Pages/2010.html#30)

The complex portion, the part that we cant do from aluminum, is two carbon halves bonded together then that cone is bonded onto a aluminum plenum that was cut in half. an aluminum flange is then bonded to the top for the restricter/TB the carbon halves were made from two female molds we cut from MDF on a CNC router we have access to. all wet lay up as well.

[carlosm]

http://arizonafsae.com/Univers...y/Pages/2010.html#30 (http://arizonafsae.com/University_of_Arizona_FSAE/Gallery/Pages/2010.html#30)

made from carbon for pretty cheap as it is all wet layup and we did it ourselves, works great! the plenum and runners are aluminum. the molds were made on a CNC router from MDF and have yielded some nice results. (except for me putting a piece of carbon on all crooked. You may want to reconsider the carbon option, its the best way to get complex shapes.

[Gaanja]

@Wesley: could i get a look at your models with and without the fillets?? I could get a better idea of what you mean.

Btw when we ran the Ansys tests, there was a 4 percent variation of pressure distribution in the outer runner entries with the box type sharp contours. With the fillets, I got a distribution of 26% each in the inner runners and 24% in the outer. It was 30 % on the inner and 20% on the outer runners without the fillets. The

[Wesley]

Ganjaa, what kind of simulation did you run? As in how did you determine the variation?

[Gaanja]

We did steady state analysis by specifying inlet velocity and outlet pressure. Iterations were done over many sets of inlet and outlet parameters. The variation was determined as the percentage of mass flow rate at each runner outlet to the total mass flow rate at the inlet.

[Wesley]

So you did this with all of the inlets open at the same time?

[Gaanja]

Yes.. The results when keeping just one of the runners open resulted in a net drop in mass flow in the outer runner

[Wesley]

Yeah, even flow distribution is pretty difficult with a four cylinder, that's why an ecu with cylinder trim is pretty useful.

Well, I'm not going to argue against your test results since I don't have any hard data with me.

You could consider hand forming the fillets and welding together, but that takes more expertise. Do you know any shops that do body work? Some of them should be able to form the metal by hand.

I also don't have pictures of the setup I'm talking about, I'm currently away from my home computer. But take the fillets off of what you have now and that's basically what I mean. Just make four panels, top, bottom, and strips for the side, and weld those together

[Gaanja]

Thanks a ton mate... really appreciate it.

This how i thought i could get fillets done on sheet metal. We god wooden patterns done in the shape of the fluid that fills the plenum and as a shell. We had asked the guy to use 3mm thick aluminum sgeets in 6 series. I guess that was too hard for him to work on. I sent 2 mm thick 1 series alu. My only worry is that 2 mm might turn out to be a little too thin and the plenum might breathe when the engine is running, considering the flat surfaces its got. I should run an FEA analysis and figure if it will flex but the guys doin FEA are really too busy now.

I'd just like to get your thoughts on if you think its a feasible option.

cheers

[Chris Boyden]

I built a aluminum plenum in a triangle shape way back in 2004.
The base of the plenum where the runners came in was 1/4" plate (~6mm) 2" wide or ~5cm and was a little longer than all 4 intake ports on the head.
The main body of the plenum was sheet 6061 T6 0.049" thick ~1.24mm.
The triangle tapered toward the restrictor. The height of the triangle was ~10" or 25 cm..

Anyway...this will give you an idea of the area that was exposed to engine vacuum.

This intake worked for a while... how many hours, I can't remember. However after revving the engine
and closing the throttle which created a high vacuum, the intake was "breathing" heavily.
The cyclic vacuum loading eventually caused fatigue cracks on the plenum, and it split it open in short order before competition.
We welded the cracks which allowed us to finish competition.

Later in the summer it split again, and we rebuilt it using 0.063 or 1.6mm aluminum sheet 6061 T6.
The difference in stiffness was considerable. The "breathing" or flex under vacuum was miniscule (not measured),
but it never broke again. So, I would have to agree with the other posters that 2mm material is probably sufficient.

[Wesley]

I built 2008's from 0.05, and it stood up really well, of course the flat sections had a rolling curve in them for added stiffness (which actually happened because I had to cut and reweld it to clear the rear envelope, but it worked out well.)

[Gaanja]

We had a similar contour for the 09 car. I was hell bent on maintaining the inner contour properly to keep the flow behavior as close to the cfx simulation as possible. I got it done as a 2 piece sand casting splitting it by the vertical mid-plane. It all went fine till here. We got the inner surface buffed and that coughed up a few visible blow holes. We filled it up with epoxy and then buffed it again. But when we got the 2 pieces welded up together the epoxy burnt away. I guess what i should have done was fill the blow holes with weld beads and buffed the inside again.

But nevertheless i can say that the intake performed beautifully. The response was much better than any other year and it gave substantial increase in horse power and torque than previous years. Also, we have a motec M800 and i didn use any individual cylinder trim. i can say the distribution was even because the color at the tips of the spark pugs were even. If the outer cylinders were running rich the tips should have been more sooty.

Considering we have very little time left for Italy '10 i dunno whether its a good idea to get it casted again. The team expressed great displeasure over the weight concerns last year.

[dazz]

http://sphotos.ak.fbcdn.net/hphotos-ak-snc3/hs358.snc3/29504_396789816967_547636967_4651443_5151747_n.jpg
This is a concept for a new plenum for DRT's single, eliminating the need for internal bell mouths. Just a quick paper mock up to see how it looks. The idea is to tweek the design before making one out of sheet metal, smoothing out the low spots with bog ready to take moulds from.

Then we can either make it in two halves, or go with the solid wax method mentioned erlier.