Anyone know how to design a gasketed joint?

[Josh]

Am putzing with the blowers again, and, am trying to make them a bit more user friendly to install and work on this time around. There were two reasons I took them off last time: The belt jackshaft drive mechanism, and, the installation and removal of the blowers themselves. I've figured out a better way to drive them (was very simple, actually. I was making it much more painful than it needed to be), but, I need to figure out how to mount them.

Originally, I had intended on using the (4) 6mm bolts that held what was the discharge cap on the blower when it was in the Ford Supercoupe. However, those 4 bolts are on the gasketed surface I am tring to clamp to. So, In order to clamp them using those 4 bolts, I would have to have a intake design where I can get to the heads of those 4 bolts inside the intake manifold. That was the problem I ran into before. I ened up trying this messy thing with long, long bolts, compression sleeves, etc, didnt turn out so hot. The heads of those 4 bolts were right by the valve covers and hard to remove... not a good option.

The blower was mounted to the supercoupe in the OEM application using 3 large 12mm bolts. The are arranged in a triangle, one on each side in the back, and one in the front. My idea now is to use these 3 bolts to clamp the blower to the manifold. Basically, pinch the blower tight against the intake manifold on its gasket surface. What Im not sure of is if I will be able to get enough clamp load, and, if I do clamp it down hard enough to get a good seal, if it will flex the case of the blower, which, given the precision nature of the beasts, would be very, very bad.

I've done some hunting, and cannot find any specific walk throughs on how to design a gasketed joint (could go buy a college textbook on it, but, I'm trying to avoid reading 400 pages and spending 100 bucks on a book). I figure you have the following factors:

Clamping force of the bolts
weight of the blowers
moment of the belt on the end
surface area of the joint
type of gasket material
blow out pressure (pressure pushing on the side of the gasket)
pressure of the contained substance (in this case, air @ up to 25 PSI)
temperature rance (0 to 350* in my case. The blowers get VERY hot under load)
surface area of the air discharge port
Thermal expansion of the materials changing the clamp load value


If I use (3) grade 8 7/16" bolts, I get 16KSI of clamp load per bolt max. That gives me a max total clamp load of 48KSI

My air discharge port area is 4.5X5.25 inches. That's 23.625 square inches, multiplied by 25 PSI, gives me 600 Lbs of force from the air trying to push the blower off.

The blowers weigh 40 lbs each, they will be hanging off the side vertically, so, it would be a complex mathematical equation to figure out what % of total weight is trying to separate the joint, so I'll just use 100%. That gives me 640 lbs trying to separate the jont.

My gasket area is .6" all the way around, calculate it out, and you get 13.4 Sq. In. of contact area.

If I go with a 1/16" thickness gasket, my blow out pressure is 25 PSI times .0625 inches times a parameter of 20 inches gives me 30 lbs of force. That brings me up to 670 lbs.

Nxet is pressure rise due to temperature. Using the P/T=P/T relationship, I can figure that the rise in temp with (theoretically) drive the pressure up 25 PSI. Clearly, thats not right... Let's figure the temp change drives 5 PSI (probably still 3X more than reality) thats 5 times 23.6. or about 125 Lbs. That gives me 800 lbs for force.

Moment on the end of the blower from the belt will be in plane with the joint, meaning that it will be trying to shear the joint, not a compression/tension force. Not sure how to calculate this in, we'll ignore it for now.

so, that gives me a worse case of 800 bs separating force. McMaster-Carr sells 400* max temp gasoline safe Carbon/Buna-N gasket material with a 2000 PSI max clamping force. 2000 PSI times my contact area of 13.5 Sq. In gets me to a total force of 27KSI.

so, assuming a factor of safety of 2 (I know I am already high on some of my variables, like the 25 PSI of boost pressure and the pressure due to temp rise), let's assume a max force of 1600 lbs trying to separate the joint. Divide that over my contact area of 13.5 Sq. In, and I get 120 PSI. That leaves me with over 1600 PSI of clamp load to hold the blowers in place, resist vibration, belt moment, etc. I think that I could easily use the 3 large 7/16" bolts to clamp the blowers down, and, could possibly even step down to a grade 5 bolt to save a bit of cash (these will need to be long bolts, in the neighborhood of 5-6" so grade 8 ones will be spendy).

The last questions left are: A) How the blower case will react to that kind of torque being put into it, and B) what happens to the joint when everything gets hot. I would think the best place to start would be to take a look at the torque spec for those from Ford, as applied to the Supercoupe. Unfortunately, trying to find this seems to be diffcult... Searched for a while and came up with nothing. however, a 12mm bolt is a pretty big bolt, so, I am assuming that I can't put enough torque into them to warp the case. The bosses on the side the bolts pass through are HUGE. about 1" square and 2" tall. They look to be designed to spread the load through the case pretty evenly.

Last, but certainly not least is thermal expansion. I want a max 2000 PSI clamp load. aluminum expands at .000013 in/in/*F. Steel is .0000072 in/in/*F So, a 350 degree change over a 6" tall aluminum case will be about 27 thousands of case growth. The steel bolt will grow 15 thousandths. That means that the thermal expansion will apply 12 thou of stretch to the bolt. I hunted for a chart on what 12 thou of stretch equates to in PSI for a 7/16 bolt and didnt find anything. I know they exist, I remember doing them from tables in books back at school... back to that textbook thing. Anyway, there are 20 threads per inch on a 7/16 UNF, that means the bolt gets pulled .05" for each rotation of the bolt. My joint needs 27KSI of clamp load to acheive max PSI load on the gasket. 27KSI divided by 3 bolts is 9KSI per bolt. Each bolt generates 16KSI at 70 Ft-Lb, max load. Since I'm operating in the elastic region of the bolt, I want to torque it to 56% of the bolt's working range (9KSI divided by 16KSI), or, 40 Ft-Lb. 12 thou divided by .05 inches per revolution is .25, or, 1/4 of a turn. So, I'll have to torque the bolt down to 40 Ft-Lb, record the number of turns, then back it off 1/4 turn and see what the torque value is.

So, does anyone see anything I missed, or miscalculated? As I see it, I think it will work On to designing the intake...

[M37UK]

Blimey Josh, I dont know about the rest of the guys here but are you honestly asking us to confirm THOSE calculations ??
I think its a rocket scientist you need to ask.

Cheers

[Tanner]

I like pie...

'Tanner'

[Master Yota]

Cripes Man, you won't buy a text book but you'll write one?

I'm going to pretent that I understood all that mumbo-mathematical-jumbo, and to an extent I did, but correct me if I'm wrong, but did you account for two blowers pushing pressure into one intake? Would you not have to make your calculations at 50psi instead of 25psi? I would imagine that you could control the amount of pressure with a blow-off valve, so that you wouldn't risk blowing the gaskets out. Does the air pressure comming from the turbo's (already higher than atmospheric pressure) affect the pressure produced by the blowers? (for example - does 10psi output from the turbo into the blower increase the ouput from the blower by 10psi - {25psi at atmospheric pressure to 35psi via force feeding} ?)

I would make the assumption that Fords engineers have already designed the 3 point mounting system to be not only overkill for the pressure application, but reliable for long term usage as well. I would use Fords mounting system, and then change if it starts blowing gaskets out. As for the 7/16 gr8. bolts being spendy, I buy them all the time for about 1.50 each. Considering the amount of money already in the truck, I think you could afford the extra $4.50

[Josh]

the snails will raise the blower pressure, but, here is where it gets weird. The snails are compressive, meaning they pressurize the air as it passes through, while the screws are volumetric, meaning they simply move whatever pressure is applied to them, and rely on the backpressure of the downstream system to provide the resistance to raise pressure. The blowers put out 90 CID per rev, and the engine takes in 100/rev. It's 400 CID total, but, I am splitting it into 1/2, since there are two blowers, then, 1/2 again, since it is a four stroke, which requires 2 rotations to swallow the complete swept volume. My blower drive ration is 1.53:1 blower to crank, so, for every rev of the crank, I get 1.53 revs on the blowers (assuming no belt slip). Other things to factor in are VE and temperature, but, to keep things simple, I'll leave those out. I have a huge spreadsheet I've put together that factors those things in for the turbos, blowers and engine. At a 1.53:1 ratio, I take in 100 CID at the engine, and get 138 CID out of the blowers. That gives me a pressure ratio of 1.38, or 1.38 times whatever is put into the back of the blower. If I get 10 PSI out of the turbos, that is 1.67 times atmospheric, assmuing atmospheric of 14.7 PSIA. To get your final pressure, you multiply the two ratios, so, 1.67 times 1.38, and you get a final ratio of 2.3. subtract out 1, for atmospheric, and you get 1.3, or, about 19 PSI of boost. In reality, it will be higher, due to temperature, or volumetric efficiency.

The interesting thing in all this though is the throttle response and turbo lag. Typically, with the size turbos I put on there, I would get full turbo boost at around 33-3500 RPM, the throttle response would be so-so, and it would be laggy. Since I will have the blowers driving mass flow up at lower engine speeds, as well as thermal efficiency, I should get full pressure from the system at as low as 1800-2000 RPM! The other big thing is that you can see it doesnt take much of a pressure change at the turbos to make a very large pressure change at the piston. If I went to 15 PSI on the turbos, I get 2.0 times 1.38, which is a PR of 1.76, or 25 PSI of boost. Teh last big benefit is that by force feeding it with the blower, I make the intake manifold the highest pressure region. Typically on a turbo engine, the exhaust manifold is the highest pressure region, which means you have to run short overlap cams, with medium durations and short lift to minimize overlap. If you get overlap on a turbo engine under load, teh exhaust will actually back up into the cylinder. By runnign the screws at 5 PSI, I generate just enough pressure to push the exhaust out with a high duration long overlap cam, which will help keep the cylinders cool, and push the efficiency up even more. This is why most WW2 aircraft engines are twincharged. It wasn't that the turbo wasn't big enough, the screws were used a a blower to scavenge the cylinder and push the hot exhaust gasses out, keeping everything cooler, letting them run the engine harder.

Ford had one very big key difference in their mounting Vs mine. In their mounting, the bosses of teh blower sat on bosses o the intake, meaning all of the compressive force traveled through the boss into the intake. On mine, the foce travels through the bolt, into the boss, through the case, down to the gasket surface. that will put strain into the case of the blower. Not sure if that is something I want to try yet or not, as that seems like a bad idea when two rotors are whipping past one another by mere thousands of an inch at 10K rpm, and then there is thermal expansion to factor in...

[Master Yota]

Looking at the blower cases in your Sig, they appear to be quite a rugged setup, and made from aluminum(?) I wouldn't worry too much about the case flexing. Could you redesign your lower plenum case out of thicker material to prevent distortion? Or figure out how to weld some cooling fins in to help disipate the extra heat? An intercooler on the turbo's might also help, as the incomming air charge would be cooler, naturally subtracting from the overall heat that the blowers would be generating...? Is it worth considering a method to water cool your plenum case, or design it with an air gap to help cool the charge before it gets to the manifold?

[Josh]

Ray tossing all those ideas around. A lot of it will be dictated by packaging size. I would love to fit an intercooler between the screws and the intake. There's not enough room for one between the turbos and screws, I know that much fo sure already, but, I plan on spraying methanol/water into the airstream upstream of the screws. the meth/water will help keep the screws cool and provide a hydrodynamic seal, so, that should boost blower VE. The problem with spraying meth above the screws and having an intercooler downstream is that if you get a backfire, it's going to make a mess. the throttles will be between the snails and screws, so, there would be nothing to keep the fire from traveling back into the intercoolers and blowing them out. I could run a 75/25 water/meth mix, that ratio will not ignite without being put under some serious pressure.

[Master Yota]

I was leaning more towards an intercooler between the turbo's and the screws, plumbed down infront of the rad or somewhere else viable... Another option, and I hate to suggest it, but a low profile hood scoop, with an electric fan mounted underneath to blow cooler air over the top of the setup might be enough to keep the temperatures happy...

[Josh]

yeah, no hood scoop, lol. that was one of my original design criteria, it had to fit under the hood, and above the bottom of the frame. I gotta take some mesurements and see what kinf od room I have to work with.

[Master Yota]

Could you duct air from the cowl vent to blow onto the back of the screws? Another option would be to redesign the side curtains under the hood to channel extra air into the engine compartment...

[SOTVEN]

I like pie...

'Tanner'

ME TOO!!!! LOL!!!

[SOTVEN]

Blimey Josh, I dont know about the rest of the guys here but are you honestly asking us to confirm THOSE calculations ??
I think its a rocket scientist you need to ask.

Cheers

DITTO THAT

[Josh]

Could you duct air from the cowl vent to blow onto the back of the screws? Another option would be to redesign the side curtains under the hood to channel extra air into the engine compartment...

Hmm, those are good ideas. Might have to think more on those!

[poor rich]

Well Josh,
???? it ??seams... ummm??
Ok try this, they say a doctor makes the worst patient, and the last person you ever want to do mechanical work for is an engineer.
sometimes, you have to accept ( or hope ) the engineers before you knew what they were doing.
use the same type bolts and gasket material Ford did, chances are that will work.
Years ago NASA spent $600/ "pad", then discovered the off the shelf Depends worked better for about 1/600th the price.
Also many here will tell you K.I.S.S. Keep it simple stupid.
Your friend and life adviser.
Rich

[Frank Irons]

Why don't you use a copper O ring or Gasket?

Frank