Sharing the load- friction and dowel pins 1

[BobM3]

I'm going to bolt a bearing block to a gear housing. The block will see a thrust load. I can't provide enough bolts to generate the clamp load needed to ensure that friction between the block and housing will react the thrust load. So I'm thinking of adding some dowel pins to help. The dowel pins will provide some shearing reaction force. I'm struggling with the load sharing between the two. I'm afraid the friction will react all of the load until there is slippage and then the pins will react all the load. When that happens the bearing block will have moved some amount. Does anyone have any experience with joints like these? The bearing block and housing are both aluminum.

 

[proinwv2]

From your brief explanation, what you say sounds plausible. Your design needs to be such that the load is shared.

Redesign is cheaper than repair.

Paul Ostand

http://www.ostand.com

 

[GregLocock]

Bigger or more bolts are the best solution. Your fears about how the load will split tally exactly with my experience. Dowels are reliably used for location before the bolts are snug, so the bolts can (and should) use clearance holes.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SomeYahoo]

Would it be possible to change the surface roughness to increase the coefficient of friction? Perhaps even a toothed surface?

 

[GregLocock]

Doesn't really work. At a micro level all surfaces are pyramidal.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[davidh729]

Can you design the joint to react all of the load with the pins and just hole the poieces in place with a coiuple of bolts?



Dave Hyman
iRobot Corp

http://www.irobot.com

 

[TheBlacksmith]

My experience also says it will fail. Generally speaking, the "pins" have to be designed to take the load and carry the full load. Same failure mechanism has occured with bolts and welds; the weld resists all movement, precluding loading the bolts until the weld yeilds, then the bolts are overloaded and the joint fails. Should be all friction or all dowels. Look at more bolts, or larger bolt(s) to attain needed clamping force.

 

[CanuckMiner]

TheBlacksmith is right. I was recently involved in an investigation into a pillow block failure, in part due to the lack of sufficient clamping force from the bolts.

Once the block slips, how can you reliably estimate how much "kinetic" friction is present? The answer is that you can't. Therefore, the safest design is to provide a fixed shoulder to resist the block movement. In our case, we welded stop blocks in place, designed to withstand the entire side load.

Cheers,
CanuckMiner

 

[DAVIDSTECKER]

Is th reason you can not generate enough clammping force because you are using threads tapped into the housing and it is a weaker material than a grade 8 nut?
If so can you work the design to incorperate a nut?
If you have a nut and bolt arangement and are using grade 8 you might see what an "L-9" nut and bolt will give you for clamping force.

 

[Qshake]

I must be missing something but as I see it once the friction fails the bols and shear pins will share the load.

The bolts themselves are always checked for the ultimate condition of friction failure and transfer by shearing in structural connections. THe pins can only add to this connection.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[Pressed]

As all the variables are unseen I can only suggest that you mill a tight fixed pocket to install the block into. That way, if I understand the setup, the bolts would have to break or the block shear in half for movement to occur.

Or mill a looser pocket and use shims and hope they stay in place.

_______________________________________
Feeling frisky.........

http://www.tailofthedragon.com

 

[desertfox]

Hi BobM3

In addition to the bolts and dowels would it be possible to fit a keyway to help take the shear load.
The other concern I have is that you do not give any details of the block or the position of the thrust relative
to the mounting bolts.
If the thrust is offset from the mounting bolts then in addition to the shear load you might be faced with bending loads as well.
Perphaps you could provide more detail for us to assist you.

Regards
Desertfox

 

[BobM3]

Thanks for all the input. You've confirmed that mixing the two is not a good idea. I like bolts and friction but I need location also. Dowel pins are a cheap way to locate (this will be a production gearbox so we're trying to keep costs low). There is also a radial load so I need the bolts to hold the block down.

There's not enough room for all the bolts I would need to get the friction nor would increasing the grade of the bolt be enough. I can't allow the bearing block to move as it has to locate gears on a shaft. Bending loads at the joint will occur but I have a handle on that.

Milling pockets or shimming or using stop blocks is an option at this point. Of course any way I do it will have to be very stiff to be able to share the load with the friction joint. I can't think of anything more stiff than the friction load path.

 

[RCarman]

Actually, this basic idea is commonly used on very large (~80 ton) machine tools to retain drive gears (racks and circular sectors) for high output feed drives. The difference is they use tapered pins, the holes for which are reamed at assembly, so there is no clearance to give problems. The bolts mainly function to hold the parts on the pins. Obviously the shear strength of the pins must be sufficient to carry the load.

Properly done, this arrangment works great, and there are hundreds if not thousands of successful machines out there using it, many of them in service for decades.

Typical practice is to ream the holes shallow in the factory, which allows some re-positioning (via re-reaming) in the field for final alignment.

The tapered pins are a little special, in that they have an internal thread at the large end of the pin to facilitate removal. The taper angle is selected to be self-locking, so no other retention means is necessary.

 

[BobM3]

Interesting! In my case, I need to also take into account the cost of assembly so I'm not sure I can have the assemblers ream at assembly but I like the idea of the self locking tapered pins (so no bolts needed to hold the bearing block).

 

[RCarman]

What I meant was that no other retention means was necessary for the pins. (A common design for tapered pins is to have a thread on the small end for retention.)

We still use bolts to hold the item in question (i.e. gear rack) on the pins. This is to eliminate the risk that a lateral force could knock the pins out.

 

[Tmoose]

What is the sequence of the processes for establishing the block location, and the subsequent gear alignment?

I'm having some difficulty imagining a bearing block that can't be bolted securely against thrust loads. A full sized feature via a machined block resting in a register on the main housing is the way to handle emergencies. Dowels Fitted to ensure alignment are fine, and even sized as duplicate system to handle loads. Varying loads mean Wiggling parts, even if only on a micro-wiggle scale. Wiggling parts restrained primarily by Dowels will fret and wear the mating faces, dowels, and dowel holes and someday be loose. Ask any Aircooled VW racer about their doweled crankshaft and flywheel. They will describe with pride that they doubled the number of crank/flywheel dowels from 4 to 8, using a fancy fixture to attempt to create 16 coaxial holes.

http://www.actionimport.com/images/EMPI/EMPI-00-4099-0-L.jpg.

If their motor was so powerful those holes get beat up, they fit oversize line-reamed dowels, or really stepped up and payed to have a taper OD fit added, or simply bought a crankshaft with a proper bolted flange.

http://www.vw-store.com/Images/Scat Flange Crank.jpg

VW and Porsche abandoned dowels as a significant flywheel attachment detail in the 60s, as soon as their engines exceeded 100 HP or 4 cylinders.

Ask any Chevy or Ford or Dodge racer about what they had to do to upgrade their 6 or 8 bolted flywheel attachment, and they may have bought some better bolts for $20.

Is the block really a cap, with the bearing bore half in the block, and half in the main housing?

http://www.4x4wire.com/toyota/tech/detroit_locker/open_diff2.jpg

Even if my calculator says the shear from the thrust can be handled by dowels, I keep picturing a significant moment being generated too, trying to pry the block off the housing, and the bolt preload had better be up to the task.

 

[WhiteTiger]

Where load must be shared between pins and bolts of limited numbers due to space constraints, the solution is often use of tapered bolt holes in the mounting and split taper seats under the bolt head to ensure that the bolts themselves have an initial significant load share.

Doesn't always apply, but in marginal situations it can make the critical difference.

 

[BobM3]

Tmoose -

My aluminum block is similar to a pillow block bearing. I'm reacting loads from worm gearing - lot's of thrust on the shaft. Yes, there will be overturning moments on the joint but I can handle that with clamp load. I don't like the idea of dowel pins handling the reversing loads either(especially in aluminum).

White Tiger -

I don't know what you mean by "split taper seats". Can you expand on that a bit (or is there a sketch somewhere I could look at)?

 

[WhiteTiger]

It's just an axially bored cone, center bore clearance fit on the bolt shank, with a single longitudinal slit to allow it to close down on the bolt shank as the bolt head presses it into it's matching taper seat in the mounting flange.

In effect, the tightened bolt becomes a defacto anchoring pin for the unit it's mounting.

Any comprehensive screw an bolt supplier will have them available.