Flywheel thinking... what are the dowels doing?

[Jon]

This is a really hard thread to start, because the answer is so automatic and given.
The obvious job the dowels has is to keep the crank from spinning when you are tightening the flywheel bolt.
They also centers the flywheel on the crank.
But from there on? They will obviously come in to play if the clamping force between crank and flywheel was overcome.
Once the flywheel is actually loading them with sheering forces the game is up, since the flywheel has lost it's grip.

From what I read in engineering circles a dowel is only supposed to locate things, and newer take any actual load, but it is recognized that they may help.
Of course we know that they do make a difference. I guess our practice came from what Porsche did on the 356, since that had eight dowels, but what were porsche trying to solve?
They were usually right on the money when things were engineered, would be strange if they were shooting for "may improve" in this matter.

Does anyone have some input on this?

     

[Peter]

Hello,
i think the dowels increase the friction between the flywheel and crank..
two clamped materials that are rough at the surface will slide over each other more difficultly than smooth surfaces.
if you would have no dowels at all the flywheel would spin on the crank no matter how much you would clamp it, no?
stuff to think about

[Frallan]

I really like your topic and I like the guts it takes to bring it up.
I think you are pretty much on target in questioning it.

That is why I like the high tensile "shoulder bolts" for TIV flywheels which are reamed to fit and also acting to give higher clamping force.
Plus I did not use the extra five dowels that many do put in the TIV flywheel.
I am not sure but have a theory that you do two things wrong if you add the TIV dowels, you take away friction area between the two surfaces.
You weaken the end of the crank too much and you could jeopardize the ability of the crank to hold the bolts. Very little meat left on that end after drilling those holes.

Here is a picture of a doweled TIV crank/flywheel.




As for the T1 it is similar bit at the same time, we have 40 years of eight doweling but at the same time we almost always added a better bolt and torqued 100 zillion times more. What of those two helped the most?

[dragvw2180]

 Wedgemating,               Mike

[mr horsepower]

What is TIV?

Most big v 8 use 5 bolts to connect the flywheel to the crank but they are futher from the centre  i never see one breake and you cant say that they dont have much torque.
i think bigger bolts and futher from the centre should do the trick ,

gr henri

[Trond Dahl]

What is TIV?

Type 4

[Airspeed]

I really like your topic and I like the guts it takes to bring it up.
I think you are pretty much on target in questioning it.

Couldn't agree more!

Plus I did not use the extra five dowels that many do put in the TIV flywheel.
I am not sure but have a theory that you do two things wrong if you add the TIV dowels, you take away friction area between the two surfaces.

Me neither!
Thats my feeling as well, so far just the 5 bolts have holded 500Nm 400 hp drag racing on slicks with a wheelie from a 1000kg super...

As for the T1 it is similar bit at the same time, we have 40 years of eight doweling but at the same time we almost always added a better bolt and torqued 100 zillion times more. What of those two helped the most?

Easy: the better bolt with twice the torque load.

I believe in this so much actually, that the T1turbo I am building now, will NOT have 4 extra dowels but I will use a Berg bolt, just nt siure how much more torque I will give it.. 

[mr horsepower]

Walter buy one extra and find out what it can have or not on a old cranck and flywheel.

gr henri

[Lee.C]

Good Man JHU  A proper question! (ps diggin the MacDowells pic )

I have often wondered just how much work those extra dowells really do    Maybe help with "SHOCK"/Sheer force 

I have had Gary WEDGEMATE my SPG - he says it good for way over 300+bhp 

[Frallan]

I have had Gary WEDGEMATE my SPG - he says it good for way over 300+bhp 

Wow, you mean wedgemate between every piece of the crank or just the flywheel?
That is as advanced as the splined needle bearing crankshafts that existed in Germany in old times.

I have been involved in a friends SPG some time early 80´s when he TIG welded the pieces. It held at least for a full Swedish summer of dragrace starts.

[58vw]

shear value/force

[Bruce]

The obvious job the dowels has is to keep the crank from spinning when you are tightening the flywheel bolt.

True.  Helps get the flywheel on in the exact same location your balancer did his job.

They also centers the flywheel on the crank.

True again.

They will obviously come in to play if the clamping force between crank and flywheel was overcome.
Once the flywheel is actually loading them with sheering forces the game is up, since the flywheel has lost it's grip.

I don't agree with this.  If the fit of the pins was really loose, then the flywheel would need to rotate in order for the pins to be loaded in shear.  But a properly machined crank/flywheel has the bores for the pins match reamed.  Thus they are in shear the instant you push the flywheel onto the pins, and also when the bolt is tightened.

From what I read in engineering circles a dowel is only supposed to locate things, and newer take any actual load, but it is recognized that they may help.

That's one way a dowel pin is used, but it's not the only way.  If it was, there'd be no need for hardened pins like we use on our cranks.

There are two mechanisms for torque transmission between the crank and flywheel. 

If you bolt the flywheel on without any pins, the clamping force of the bolt creates friction between the surfaces.  It's not just the surface between the crank and flywheel.  The underside of the bolt head to the washer, and the washer to the flywheel also helps transmit the torque.
The other way is through the shear strength of the dowel pins.  8 pins are twice as strong as 4.  Hardened pins are stronger than soft ones.

Someone above mentioned wedgemating.  This is just applying a Morse taper to the parts. 
 
I believe they use #3 (correct me if it's wrong).  This taper has an angle of 1° 26' 15'' (1.4375º).  When you torque the bolt, the clamping force is magnified by the angle.  If my math is correct (applying simple high school trigonometry) this factor is 39.8.  In other words, it is like being able to torque the stock bolt that much more than stock.  So if you torque your bolt to 500 ftlbs, you get the same friction between the flywheel and crank as if you torqued the bolt to just under 20,000 ftlbs!  It's no wonder it works!!!

[Neil Davies]

Plus I did not use the extra five dowels that many do put in the TIV flywheel.
I am not sure but have a theory that you do two things wrong if you add the TIV dowels, you take away friction area between the two surfaces.
You weaken the end of the crank too much and you could jeopardize the ability of the crank to hold the bolts. Very little meat left on that end after drilling those holes.

With five bolts AND five dowels, you've got ten holes in the end of the crank and the flywheel rather than eight for a T1 crank. I have seen a TIV turbo motor break the middle out of a flywheel.

Someone above mentioned wedgemating.  This is just applying a Morse taper to the parts. 
 
I believe they use #3 (correct me if it's wrong).  This taper has an angle of 1° 26' 15'' (1.4375º).  When you torque the bolt, the clamping force is magnified by the angle.  If my math is correct (applying simple high school trigonometry) this factor is 39.8.  In other words, it is like being able to torque the stock bolt that much more than stock.  So if you torque your bolt to 500 ftlbs, you get the same friction between the flywheel and crank as if you torqued the bolt to just under 20,000 ftlbs!  It's no wonder it works!!!

I've also seen a gland nut break after many miles of use and several drag starts. With a wedgemate, could the torque applied to the bolt actually be lessened? Would it be ok to use 350ftlbs instead of 500?

[Lee.C]

I have had Gary WEDGEMATE my SPG - he says it good for way over 300+bhp 

Wow, you mean wedgemate between every piece of the crank or just the flywheel?
That is as advanced as the splined needle bearing crankshafts that existed in Germany in old times.

I have been involved in a friends SPG some time early 80´s when he TIG welded the pieces. It held at least for a full Swedish summer of dragrace starts.

Na just the fly wheel - The SECTION'S are "pinned" and then "plug welded" This is all removable if you ever need to rebuild the crank in the future 

[Frallan]

Quote Neil Davies:With five bolts AND five dowels, you've got ten holes in the end of the crank and the flywheel rather than eight for a T1 crank. I have seen a TIV turbo motor break the middle out of a flywheel.
:end quote
...................................
Please look at this picture as it is more visiible on the crankshaft end. Put a large dowel (compare to the TIV flywheel picture) in between every bolt and the material left in the crankshaft end and the area for the flywheel to attach to the crank, is not a lot.

A German company makes 12,9 strength bolts with the nice design that they are done with a shoulder that is oversize.
In order to fit, you have to ream the flywheel hole exactly for a very tight fit when assembling.
This is what I prefer and it leaves the material in the crankshaft.

Now, when we really are free to design, we do it like Scat, Pauter and other billets with a large flange and more bolts.

Now I am not sure all those bolts are needed.
Look at the V8 racing 180 degree crankhaft. No dowels and only a wide flange with few bolts.
It only has six.



Let the brainstorming go on.

[Lil Gasser]

Quote Neil Davies:With five bolts AND five dowels, you've got ten holes in the end of the crank and the flywheel rather than eight for a T1 crank. I have seen a TIV turbo motor break the middle out of a flywheel.
:end quote
...................................
Please look at this picture as it is more visiible on the crankshaft end. Put a large dowel (compare to the TIV flywheel picture) in between every bolt and the material left in the crankshaft end and the area for the flywheel to attach to the crank, is not a lot.

A German company makes 12,9 strength bolts with the nice design that they are done with a shoulder that is oversize.
In order to fit, you have to ream the flywheel hole exactly for a very tight fit when assembling.
This is what I prefer and it leaves the material in the crankshaft.

Now, when we really are free to design, we do it like Scat, Pauter and other billets with a large flange and more bolts.

Now I am not sure all those bolts are needed.
Look at the V8 racing 180 degree crankhaft. No dowels and only a wide flange with few bolts.
It only has six.



Let the brainstorming go on.

Scat went to only 1 locating dowel and just the bolts on there flange crank a few years ago now,I wonder why?

LG     

[Jon]

I really like your topic and I like the guts it takes to bring it up.
I think you are pretty much on target in questioning it.

Thanks Frallan,
It seems the only thing harder than to start this thread is to continue it.

In the world of machine design a bolted flange joint is a very common occurrence. And it's quite easily calculated, basically what you do is to divide the load you want to transmit on the bolt quality/size you would like. This gives you a number of bolts you need. (Very simplified). These are easily calculated and the formulas are old as the hills, however there are no common practice for calculating how much load a dowel can transmit, and engineers looks like question marks when you ask them. There is a reason for this. It depends on a whole range of things, what material the holes are drilled in and also the size, insertion and material of the dowel itself

What we are doing is something completely different, we are asking the original solution to withstand significantly more than it first was designed for.
To add to the challenges we have a bolt (I use the term lightly as it's actually more a tube) designed to clamp the flywheel with about 130 nm. And it's not even designed to stretch like a proper bolt commonly used in a dynamic environment.
It will stretch of course but most lightly it will stretch in the thinnest part of bolt. And that is the last 2 mm before the threads, thereby putting extreme tension in an small part of the total lenght.
As soon as VW started to design engines with higher output they went directly for five proper tension bolts on the type 4. Much better than one "stiff" bolt on so many levels. Just think about how much force is lost under one big bolt head.

About going BAS or not...  it comes down to if you believe that dowels are better than a stretch bolt that keeps the clamping force. Also remember that the nature of the BAS puts ALL the stretch in the last thread just as the stock VW bolt. It's just bad design. I would throw them BAS bolts as far as I could and buy correctly designed 12.9 bolts or ARP 15.9 bolts.
 
This was what I remember from bolt design at school, mixed with my 0.2 cent.
More about the dowels later...  it's time for lunch now!

Updated 12:30

[Jon]

They will obviously come in to play if the clamping force between crank and flywheel was overcome.
Once the flywheel is actually loading them with sheering forces the game is up, since the flywheel has lost it's grip.

I don't agree with this.  If the fit of the pins was really loose, then the flywheel would need to rotate in order for the pins to be loaded in shear.  But a properly machined crank/flywheel has the bores for the pins match reamed.  Thus they are in shear the instant you push the flywheel onto the pins, and also when the bolt is tightened.

When you are drilling and reaming the holes there are no shearing forces. When you insert the dowels there are no shear forces, as both parts are stationary. Once you are tightening the bolt, the holes in the flywheel is resting on the left side of the dowels. When fully clamped this microscopic shear is put to rest, and you can regard flywheel and crank as one, until the clamping force is overcome. At that instance the flywheel is moving to the left and it doesn't stop until it reaches the right side of the dowel. At this point the bolt is slightly un-tightened and the flywheel is resting in a new location, this is of course on a microscopic level. At this point the dowel pins may or may not save the day.

From what I read in engineering circles a dowel is only supposed to locate things, and newer take any actual load, but it is recognized that they may help.

That's one way a dowel pin is used, but it's not the only way.  If it was, there'd be no need for hardened pins like we use on our cranks.

They may be hardened to work satisfactory when dealing whit heavy parts. Where have you learned the part about using them to withstand sharing forces.

If you bolt the flywheel on without any pins, the clamping force of the bolt creates friction between the surfaces.  It's not just the surface between the crank and flywheel.  The underside of the bolt head to the washer, and the washer to the flywheel also helps transmit the torque.

The forces you claim the bolt head is seeing are the same it would need to unwind... let's hope you are wrong.  If there was a reason to add this into the calculations I bet that you would find formulas for this also. 

[Lightning]

Trying to calculate the actual transferable torque of the flywheel to crank connection is not straightforward. Still, I have given it a shot:

Starting with the 'easy' one, the type 4, because this is a standard 5 bolt flange connection. By using fairly conservative factors for friction rates in threads, under the bolt heads and between the mating surfaces of flywheel and crank I get values between 780 and 1100Nm. Not many have engines which can produce this kind of torque, so the flywheel should always stay in place providing the assembly is done properly.

For the type 1, it's a bit more tricky since the center bolt is not really a bolt. Anyway, considering it to be a normal bolt and applying the same method and factors gives values of 200-250Nm. This is in a range where tuned engines operate. The dowel pins are ignored in this calculation. If the flywheel assembly was not done properly (dry bolt, insufficient torque, ...) this value would be lower, and the flywheel would come off.

If you have had a flywheel gone off, you will have seen that the dowels have not done anything to help you. Usually, the bores in both flywheel and crank are squeezed into an oval shape, the hardened dowels are the only thing you can reuse....   

[Griebel]

"About going BAS or not...  it comes down to if you believe that dowels are better than a stretch bolt that keeps the clamping force. Also remember that the nature of the BAS puts ALL the stretch in the last thread just as the stock VW bolt. It's just bad design. I would throw them BAS bolts as far as I could and buy correctly designed 12.9 bolts or ARP 15.9 bolts."

Hi,-Did you come up with this JHU ?

[Jon]

It is what I believe yes, since I don't subscribe to the load-bearing-dowel-theory (not yet anyway). Therefor I would rather use a bolt designed to stretch.

[BeetleBug]

Maybe I'm out of line here when I ask: Is this really a problem today? A couple of years back when people still used Stage 3 and sinter style clutches - yes. But today I have a feeling that even we have learned that too much is not always good.

Don't get me started on wedgemating...

[Lids]

It isn't as simple as shear and torque, think about the revs and then the flywheel still trying to move, but the crank not wanting to.  The inertia of the bolt holes been close together make it more difficult to keep fixed to the flywheel.  The best solution is a wide diameter flange on the crank to mate to the flywheel.  YOu are always goign to have a shear force, but again this could be rduced by removing weight from the flywheel.

How is an F1 crank fixed at 12000rpm and 1000bhp!  they are light and therefore not over designed.

[Lee.C]

Maybe I'm out of line here when I ask: Is this really a problem today? A couple of years back when people still used Stage 3 and sinter style clutches - yes. But today I have a feeling that even we have learned that too much is not always good.

Don't get me started on wedgemating...

I see your point dude - but whats your issue with wedgemating???

[BeetleBug]

Maybe I'm out of line here when I ask: Is this really a problem today? A couple of years back when people still used Stage 3 and sinter style clutches - yes. But today I have a feeling that even we have learned that too much is not always good.

Don't get me started on wedgemating...

I see your point dude - but whats your issue with wedgemating???

Ok, you got me started. JHU: Please remove if you think it is off topic.

Morse cone works and have been used for ages just so it's said. But a morse cone in our configuration is like asking for trouble if you ask me. What you're trying to achieve is to get it to "hook" (wedge) at the exact same time as the flywheel is where it is supposed to be. Next is the marks that the wedge make on your flywheel. If you're lucky you do not experience tiny oil leaks due to the marks. I was not so lucky.  And the marks get worse every time you remove the flywheel = harder and harder to seal.

BB

[Jon]

It isn't as simple as shear and torque, think about the revs and then the flywheel still trying to move, but the crank not wanting to.  The inertia of the bolt holes been close together make it more difficult to keep fixed to the flywheel.  The best solution is a wide diameter flange on the crank to mate to the flywheel.  YOu are always goign to have a shear force, but again this could be rduced by removing weight from the flywheel.

How is an F1 crank fixed at 12000rpm and 1000bhp!  they are light and therefore not over designed.

A flywheel is just a energy battery, and when you run the engine on a dyno the total torque is measured. If the clamping force exerted by the bolt/bolts can withstand this amount of torque + a safety margin, the flywheel will not come loose. You are obviously right about shear forces being there at all times. They acts on all moving things, but you cant see it before something fails. All machine components are designed and calculated according to a given challenge... that's why the crank is made of steel and not plaster for instance. But at some point even things made in one piece will snap off, then its back to the drawing board.

[Lightning]

Morse cone works and have been used for ages just so it's said. But a morse cone in our configuration is like asking for trouble if you ask me. What you're trying to achieve is to get it to "hook" (wedge) at the exact same time as the flywheel is where it is supposed to be. Next is the marks that the wedge make on your flywheel. If you're lucky you do not experience tiny oil leaks due to the marks. I was not so lucky.  And the marks get worse every time you remove the flywheel = harder and harder to seal.

This is exactly the problem with a wedge mate, or more precisely a conical press fit. You can't get correct press and correct position at the same time. That would require 100% control of material properties, friction factors and jet fighter machining precision. In addition, the available space at the end of the crankshaft does not allow for a suitable cone length and the flywheel is not designed for this type of connection. A more proper cone length would be min. 55mm (=diameter of crank) and the flywheel would have to be designed with a proper hub which could cope with the pressure created.

Standard flywheels have a kind of collar where the Simmering is pressing. This collar is probably permanently deformed the first time it's bolted on, and then you get oil leaks and scuff marks on flywheel and/or crank as BB explains.

This said, based on dimensions from one wedge mated crank and flywheel, and using the same factors as in my previous calculations I get a theoretical transferable torque of app. 600Nm. So in that respect, wedge mating is better than the original flywheel connection, if transferable torque is the only goal to chase.

[Frallan]

About going BAS or not...  it comes down to if you believe that dowels are better than a stretch bolt that keeps the clamping force. Also remember that the nature of the BAS puts ALL the stretch in the last thread just as the stock VW bolt. It's just bad design. I would throw them BAS bolts as far as I could and buy correctly designed 12.9 bolts or ARP 15.9 bolts.
[/quote]

I am pretty sure my nice shoulder bolts were exactly 12.9 quality.
Why do you not like them?
Does ARP have 15.9 bolts that fit this application better?
I hope I am clear, in TIV application I prfer a good quality bolt and no dowels.

As for managing the clutch, that is key to longevity AND good Drag Racing.
Many years ago I had a Honda Dragster and tried to get the clutch to do what I wanted it to do.
I ended up joining two Honda pressure plates springs in to one clutch. Now the clutch did what I wanted it to do, hold my supercharged torque. BUT I never managed to get my driveshafts to keep together.
I even built 300M ones but gave up when they snapped.
At the time I did not undertstand the rootcause....

[Fiatdude]

AND THAT IS THE REASON WHY YOU SHOULD START WELD MATING YOUR CRANKS AND FLYWHEELS

[Frallan]

Fiatdude, I researched many crankshaft suppliers when I was going to buy my first VW crank in 1977.
Scat, Okrasa, Gene Berg, and few more.
The only professional answer I got was a 7 side letter from Gene Berg himself explaining why I should buy his crank.
I did so and got a beuty of a 78 mm welded wedgemated and counterbalanced crank.
Still have it today after many years and configuartiosn of street and drag racing.

The wedge flywheel has come apart many times and still holds good. A little emery cloth on the surface prior to assembly and it keeps dry.
Love it!

Now this thread is dowels, flanges and bolts so I guess wedgemating is close to off topic?