Flywheel fastener design, clamping force and dowel shear strength 2

[yoshimitsuspeed]

The 4AGE engine uses an 8 bolt crank and flywheel because it is the performance engine and in later years came supercharged.
The economy engines use a 6 bolt crank and flywheel. The 7AFE 6 bolt crank however is used as a stroker crank for performance builds.
Many people use 2, 3, or 6 dowels to try to make up for the 2 less bolts. It is my understanding that dowels should only be used to locate a part and ideally should see no significant force. It should be clamping force between the crank and flywheel that prevents one slipping on the other.
I recently designed a flywheel of my own and using the theory that clamping force was more important than shear force I used 6 6mm bolts in between the primary crank bolts. I would have loved to use a bigger bolt but this was the easiest solution that would clear the primary bolts. This gives slightly higher clamping force than the 8 bolt flywheel and should also give a slight improvement in rigidity and I would hope some added resistance to harmonic flexing.
I have had a surprising number of people in the automotive communities question this design. I'm not talking about home builders but people who build race engines and other aftermarket manufacturers who all say they use dowels and believe dowels are better. I believe this most likely is just because that's how it's always been done, not that it's actually better. Today one of my business partners told me pretty much the same thing and suggested I research why dowels would be better. I didn't find anything convincing in a short search. I did find a good discussion here on eng-tips on dowel shear strength and where most the posts seemed to agree that dowels should not be used for strength but other posts suggesting they could. Then I decided instead of researching general theory maybe I should just throw my design up here and get opinions on the design and theory as it applies specifically to this application.
This design uses 6 M6 SHCS counterbored to sit flush below the surface of the flywheel. The main flanged head cap screws come very close to the SHCS so this ensures they clear.
Question 1) Does anyone have any input or constructive criticism on this design in general?
Question 2) is there any way in which an 8mm or even 10mm dowel would be superior to this design?
Thanks

 

[SwinnyGG]

It also has no effect on the thrust bearing or seals as I am talking about clamp force on the mating surface. Pressure plate clamp force does have some effect on thrust bearings.

Are you sure of this? Where is the thrust bearing on this engine?

I brought this up not because of concern about the crankshaft flange cracking or failing by some other means- just remember that every additional pound of clamp load you apply between the crankshaft and flywheel is an additional pound of load, and the distortion that goes along with it, that you're putting into the crankshaft flange. If the thrust bearing lives on the back side of that flange, you are opening yourself up to a potential thrust bearing/crank walk type failure. And additional 30,000 pounds of preload force into a small flange is very significant.

Not saying that you ABSOLUTELY have this problem, just that assuming that everything is fine without evaluating what effect your design has on the rest of the system is not a good mindset to have when designing engine parts. When you're clamping things together, in any application, more force is not always the best solution.

 

[yoshimitsuspeed]

The thrust bearing is between cyls 2 and 3 so that shouldn't be an issue.
Even if it was at the rear flange I don't see how 6 more 6mm bolts would have a negative affect. If anything I wold think 12 bolts would create a more even deformation than just 6. The only way I see this having a significant distortion would be if the surfaces were uneven but even then I would expect more bolts to help even out that distortion.

 

[GregLocock]

I agree. The engines I can remember had the the thrust washer rather remote from the end of the crank, typically somewhere in the middle. The engine I know best only has half a thrust washer, but that's another topic.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[gruntguru]

What loads do thrust bearings need to carry apart from clutch throwout bearing load and perhaps thrust from a helical timing gear?

je suis charlie

 

[mohawk2]

Solid dowels can transmit power, hollow dowels are only for alignment. The best solution for big torque numbers is a keyed surface plus clamping force.
If you don't have the meat in the part to machine a keyed surface, then blind drill & fit captive solid hardened steel dowel pins then clamp together with upgraded high tensile bolts.

 

[Tmoose]

http://www.racecarparts.com/images/listings/169_l.jpg

http://i.ebayimg.com/00/s/MTIwMFgxNjAw/z/OBgAAOSw37tWBDFF/$_57.JPG

Pretty serious clutches in Top Fuel and Funny car dragracing. Once in a while there is a single dowel, but usually a bunch of bolts and 100 lb-ft + of torque get in done.

 

[SwinnyGG]

What loads do thrust bearings need to carry apart from clutch throwout bearing load and perhaps thrust from a helical timing gear?

None.

The concern is/was that additional clamp load in the flange can cause distortion, which can cause the thrust bearing face to distort, and cause the thrust bearing to fail.

Sounds weird I know, but I was once party to a project where this was determined to be the cause of a very very expensive failure.

Perhaps that experience has made me more sensitive to flange-on-bearing distortion concerns than is necessary.

 

[TugboatEng]

If you're getting this deep, why not run body fit studs? That's how the industrial side handles all of their super high torque transmissions. The only step better is a tapered coupling.

 

[pontiacjack]

Plenty of successful schemes that use only "clamp" fasteners to survive high shear forces- for instance ring gear mounting on differential cases, which transmit WAY more torque than the vehicle's flywheel.

 

[Screwman1]

Jack, there has been a lot of work on ring gears recently to add various features (hollow dowels, dowels shoulder bolts, etc. because in the higher performance versions of many vehicles they are getting slip when trying to use straight clamp. With large enough bolts and enough nut member thickness there should not be an issue, but downsizing has reduced the envelopes on many components, backing us into this corner.

 

[yoshimitsuspeed]

I still see some people saying that solid dowels would be better than more screws but no one saying why. It still seems to me that screws would be the superior method. What would I be missing?
I will acknowledge that in an application like this it would be easier to fit a larger dowel than screw but it still seems to me the added clamping force of the bolts would be better. Especially if we are talking about a slip fit between the dowel and the flywheel which would allow movement before the dowel applied any holding force.

 

[SwinnyGG]

I suspect people who are recommending dowels over screws are people who don't have any true engineering understanding of what makes dowels and screws different.

I suspect if you asked most people, even people with some technical knowledge, mechanics, etc, who have not studied, they would say that bolts work by being loaded in shear, and that a dowel is obviously stronger than a bolt, neither of which is true.

 

[Screwman1]

A locating solid dowel will not allow any lateral slip until it fails through single shear and it requires no clamp load to achieve that. A threaded fastener does not fully fill the holes (assuming that that this is a normal automotive application and the holes are not match drilled and the thread fall into the shear plane) requires clamp load to prevent lateral slip under shear loading. Once that lateral slip has occurred, then the loaded component is more likely to slip again and the fastener to loosen. In addition, the solid dowel will have a larger cross section area than the equivalent fastener that is not in a press fit hole. This is assuming that you will not be using a match drilled bolt.

 

[GregLocock]

Interestingly that post runs directly counter to the advice given in a course on joint design given by a major automotive OEM at which I scored high marks.

Ignoring clamp load in a bolted joint is like ignoring the steel in the Eiffel Tower.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Screwman1]

Gregg, If you are using match drilled dowels the joint ceases to be a bolted joint. I agree 100% that clamp load is the only way to hold a BOLTED JOINT together, but this would no longer be a bolted joint.

 

[MacGyverS2000]

I may have missed something here - but - wouldn't the bolts themselves act as dowel pins to resist the shear forces.

That was my initial thought, until it sunk in that the friction in the two matching faces offered a significant amount of resistance to movement when strongly pulled together.

With that in mind, would it not also offer better engagement to sand the surfaces rough? To move towards the ad absurdum line of that thinking, why not make the faces look more like a (flat) bevel gear? Friction fit of a rough surface plus the benefit of a dowel-like shear interface that runs the length of each tooth from face circumference to centerpoint?

Just randomly wondering...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[MikeHalloran]

I recall a u-joint flange drawing with a minimum roughness spec.
We process engineers had to intentionally break the edge of a carbide cutting tool to achieve it.
The flange was thinnish and clamping was by four bolts, and the need for additional torque capacity had been demonstrated at the test track just before production.
A proper bolted flange would have been much heavier.


Mike Halloran
Pembroke Pines, FL, USA

 

[Tmoose]

I think if a surface was rougher than (TBD) it would likely permit some embedment over time in hard service, which without routine re-torquing would lessen the bolt clamping preload, which would reduce the clamping force, which would permit micromotions, which would then let the velociraptors out of the enclosure.

 

[SwinnyGG]

Gregg, If you are using match drilled dowels the joint ceases to be a bolted joint. I agree 100% that clamp load is the only way to hold a BOLTED JOINT together, but this would no longer be a bolted joint.

He's not talking about removing all of the bolts and replacing them with dowels.


In any event, your analysis is not correct. A dowel has zero shear stress imparted to it until the assembled parts have relative movement between them, which means that the joint has already failed.

 

[GregLocock]

"would it not also offer better engagement to sand the surfaces rough"

BMW use conrods that have been snapped in two at the part line, to give absolute keying between the two surfaces. There's also several 'rough' inserts out there, which are a bit more scary as they seem to be adding compliance at the same time, rather like spring washers.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?