Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations SDETERS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Mounting Pillow Block Bearing

Status
Not open for further replies.

ldeem

Structural
Sep 2, 2013
258
I am mounting large pillow block bearings to steel structure for a bulk material conveyor. Questions have come up regarding the proper bolt torque. To calculate the bolt torque I am starting with the horizontal bearing load, dividing by the friction coefficient to determine the normal (clamping force, then calculating the bolt torque. The trouble I have is the friction coefficient. The structure if hot dip galvanized steel with large cast iron pillow block bearings. I found a value of 0.4 for cast iron on mild steel. In Marks there is 0.85 for zinc on cast iron and 0.74 for mild steel on mild steel. So it seems like 0.4 would be a fairly conservative number.
Does anyone have a “typical” friction coefficient for this application?

Secondly, for the large had pulley I am coming up with a bolt torque of 1,249 ft-lbs. I know they will need to get to this torque in stages but are there any references to how many steps they should take?

In the application the head pulley is torqued to 850 ft-lbs dry and is holding in position.
 
Replies continue below

Recommended for you

What you have is similar to a slip-critical bolted connection. There are standard analysis approaches for slip-critical connections used in aerospace, structural steel construction, etc. Here is a example of an analysis approach used for structural steel bolted connections that should work for your case. There is a reference noted in the document for faying surface characteristics, including static friction coefficient.

Hope that helps.
Terry
 
A sole plate with jack bolts that allow screws to block the bearing ends would minimize your need for a friction force to hold the bearing in place.

Have a look at this - Sole Plates, and this, Soles Plates 2. Additionally, if you could mount your bearings so that the belt reaction forces were borne by the support members, you could remove the shear on the bearing bolts.
 
tbuelna

Thank you for the reference to slip critical design. I am familiar with this approach which is similar to what I am doing. The trouble I have is the slip coefficient from cast iron bearing housing to HDG is hard to find.
 
dvd

Thank you for the reference to sole plates. The web site link looks very interesting and has a lot of useful content.

In this application I have designed the jack bolts for the full horizontal load. My goal is to design the friction force for the full load as well so in combination it will be a very robust joint.

I am curious if you normally specify machined sole plates to be tack welded to the mounting flange? I thought about this but assumed a plate would distort to the match the flatness of the mating beam flange. I have also seen many places where 1/2" sole plates have been fully welded to the beam flange. I had a situation similar to this on a large fluid coupling mount and the weld distortion was so bad we had to have a company come in and machine the base plate mounting locations.
 
[ol 1]
[li]Welded sole plates and also screwed with flathead cap screws, have been used[/li]
[li]Thick, cold-drawn plate is nicely flat and less susceptible to warping[/li]
[li]Too much welding will definitely move the flange and soleplate material[/li]
[li]Depending on the shaft stiffness, it is possible to push the beam flange around some.[/li]
[li]If these are self-aligning bearings, the bearing and the weldment/soleplate can probably reach a compromise on alignment.[/li]
[li]Taking a machining pass is good if the weldment is not too big, but for a large pulley stand, this probably won't work and can add quite a bit of expense.[/li]
[/ol]
 
In my world, for rigid stacks without gaskets, like for pillow blocks' mounting, threaded fasteners are tightened to some major fraction of yield, i.e. one value for each size and grade, without regard to the design load on each individual fastener. Basically, you don't want the pillow block to move unintentionally, even if it's used as a jacking point for some unrelated maintenance operation.

You seem to be working towards a situation where you will need to document a specific torque range for each individual mounting fastener, even if there are thousands of the same size and type.

Absent some specific contractual requirement, it seems a hideous waste of your time.


Mike Halloran
Pembroke Pines, FL, USA
 
Mike,
This is a one-off structure and there are only three bearing I need to do calculations on. So its not to time consuming. However, you have a good point if they just torqued to the usual values there would be more than sufficient clamping force.
Unfortunately, on the larger 1.5” bolts the contractor wasn’t sure they could meet the 1,950 ft-lbs torque for dry threads.
 
For big threads, take a look at SuperBolt nuts.


Mike Halloran
Pembroke Pines, FL, USA
 
I think the contractor should rent/borrow/buy a torque multiplier.

1950 lbs-ft sounds like about a grade 5 bolt.
 
Thank you everyone for your help on this issue.
 
If space permits, an 8 foot long pipe/cheater bar plus this $75 scale and I'd be good to go.

Might take a come-along etc rigged to a perpendicular location to be able to exert ~250 lb force out at the 8 foot mark.

Similarly, welding some 1" drive sockets. etc to the pipe would permit using a 250 ft-lb torque wrench according to this formula.

==========

If space at the actual installation is a problem, then one of the rigs above could be used to develop a reasonable turn-of-the-nut spec out in the open where it is convenient in about 15 minutes.

=============

The typical torque indicating washer (A325 or A490) is likely to be rated for a service a little higher than yours, but I only looked at a few suppliers on-line.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor