Compressive stress of bolts? 1

[kingnero]

I like to use four bolts for setting a heavy workbench level.
For this I would need four pieces of steel, drilled and tapped for the bolts, welded on the bottom of the feet.


Can I use (100% of) the yield stress for this (for the grade of bolt I'm using), or would this not be correct?
Now that I think of it, I've never calculated a bolt for compression. Not even a set screw.

Also, what would an appropriate factor of safety be?

I appreciate your input...

 

[desertfox]

hi Kingnero

I would not use the full yield of the bolt and include a suitable safety factor of say 2 or 3, if the threads all fail in shear and the bench drops do peoples feet get squashed?

desertfox

 

[rb1957]

bolts don't carry compression load (typically) ... compression is carried by the faces of the joint (that the bolt is compressing together).

 

[tr1ntx]

My 2 cents is in-line with desertfox & rb1957, this is a thread shear issue. What is the approximate size of bolt you're considering? How thick is the steel you're going to drill & tap? What type/grade is it? It might be weaker than the bolts.

 

[hydtools]

I would not worry about bolt compression failure. Thread shear would be the more likely failure mode. If the bench will be pushed around at any time be concerned about bending the bolt(s).

Ted

 

[rb1957]

sure this could be a shear issue, but the OP asks about compression ...

shear, of course, brings a whole bunch of different issues ... if you've got clearance holes (i suspect) then shear is carried by the joint faces (by friction), untill things move and the table bears against the bolt (which you don't want to happen so you use a preload to make sure the joint faces stay together).

this is assuming a typical bench, being bolted to a floor ... it could be being attached to a wall, then it's a different problem !?

 

[kingnero]

I'll explain a bit further: the bolts will need to carry the weight of the bench, they do not bolt the bench to the floor.
They will be used to level the workbench when it is stationary.

This workbench will weigh over 600 lbs on itself, and will roll around on 4 caster wheels. It is meant to be able to carry the weight of a car, placed on top of the bench.

When stationary, I'll unscrew the bolts a bit further out, so that the bench rests on the four bolts instead of on the wheels.

I was thinking of using M20 bolts (about 3/4"), threaded in a plate at least the same thickness of the diameter of the bolt. A flamecut 20mm sheet of S235JR, typical EU construction steel.

If the threads fail in shear, the table wil fall max. a few mm, as the wheels are still in place. No risk for personal damage, and I will be the only one using this thing anyway.

I can use the same calculations for thread failure under tension in this case, I suspect? Bolt doesn't know which way it'll fail?
If the force applied is under the max. preload it sees when tensioned using a torque wrench, thread failure will not occur?

Stationary load will be about 5 kN per bolt, dynamic load factor = 2 so 10 kN per bolt.
When a M20 (class 5.8) bolt is tensioned correctly (torqued at 200 Nm = 150 lbs-ft), it sees a preload of about 66 kN. I suppose thread failure will not occur in this case, correct?

Thx for your opinions on this...

 

[chicopee]

The original post seems to have a parallel with leveling nuts on sign and lighting post base plates. Yes, there will be compression on the threads unlike the base plates subjection the bolts to tension, compression and torsion.

 

[drawoh]

kingnero,

How long are these bolts likely to be? Will they fail by shearing the threads, or will they buckle?

JHG

 

[drawoh]

kingnero,

A couple more questions...

Would a lock nut not increase the shear strength of your threads?

Is there any reason to be working with a low FOS? Do you need low weight, low cost, or a fine thread pitch? Is there any reason you cannot just make it big and forget it?

JHG

 

[rb1957]

a typical 3/4" bolt is good for 44,000 lbs (tension, but that means the thread can handle that load in compression too) ... should be plenty for your app.

i's add the jam nut to make the bench more rigid, less likely to unwind the bolts. IMHO, no particular need to preload the bolts (not sure you can with an unloaded table ?), but the jam nut will make things secure.

reasonable question about column length ... Pcr = 10*29E6*0.0115/L^2 ... L = sqrt(10*29E6/44000) = 81" ... shouldn't be a problem !

you're using a tapped hole as a nut ... option would be a plain hole with nuts on both sides, the lower to support the bench, the upper to jam it.

 

[hydtools]

If possible, install the bolt from the top the base plate using the bolt end to bear on the floor. Directly wrench the bolt down to set it against the floor and pick up the bench load.

Ted

 

[kingnero]

@ drawoh:
How long are these bolts likely to be? Will they fail by shearing the threads, or will they buckle?
They'll stick out by about twice their diameter.
If they fail, it will be by either shearing the threads or by yielding of the bolts. They wil see minimal sideways loading.

Is there any reason to be working with a low FOS? Do you need low weight, low cost, or a fine thread pitch? Is there any reason you cannot just make it big and forget it?
I do not attempt a low FOS, I would like like to know how to calculate so that I can see in what order of magnitude my FOS (using M20 bolts) lies. If I compare the compression load to it's max. allowable tensile load, I get my FOS. I believe this to be correct.

@ rb1957:
about your preloading remarks: the bolts will be head down installed in a threaded plate, and will serve only to lift the bench from its wheels, so it cannot roll around. There is no joint as such.
I'll use a jam nut to prevent unwinding, however that will not really contribute to strength of the application, as mentioned by drawoh.

a typical 3/4" bolt is good for 44,000 lbs (tension, but that means the thread can handle that load in compression too)
That is exactly what I was trying to point out in my previous post, by saying that the bolt doens't know in which way it gets pushed/pulled.

 

[hydtools]

From this ref. compressive strength = tensile yield strength for metals

http://www.grantadesign.com/resources/materials/glossary.htm

What is your bolt material?

Ted

 

[rb1957]

any reason not to use a plain hole and support the table on a nut ... easier to replace if you strip a thread, no? how thick is the table base (that you're tapping) ?

about the preload comment ... i quickly read your post, say "preload" and thought you were trying to preload the bolt. but you're only using preload tension as an allowable, whcih is clearly conservative. FWIW, i won't use this as an allowable, as the calc is so variable; why not you the strength of the bolt (or nut) ?

why not use castors with clamps ?

how much side load can the bolt heads react ? (ie, can the bolt heads prevent the table moving sideways ?)

 

[kingnero]

hydtools: That is what I thought (see my opening post), however I wasn't a 100% sure.

as for the bolt material, I'll use a cheap class 5.8 bolt, which is a low/medium C cold worked carbon steel. Nothing special, as apparently I'll have a FOS that is larger than 10 already.

rb1957:
plates, welded on the table feet are at least 1x the bolt diameter (thus > 20mm, probably 30 or whatever i'm having lying around). They won't strip before the bolt.
why not you the strength of the bolt (or nut) ?
That is what I asked in my original post ...

why not use castors with clamps ?
Because the table/bench won't be moved around much, and this way I always have a backup at either table foot (bolt and wheel). I also don't find clamped casters as stable as fixed feet.

how much side load can the bolt heads react ?
a 600 lbs welding table won't be going anywhere soon, and there will not be great sideways forces applied to.

Main reason is to have a sturdy backup if the castors should fail with a car on top of the bench. Max. it can "fall" is a few mm, and the bolts will then be there to avoid further failure.

 

[rb1957]

"plates, welded on the table feet are at least 1x the bolt diameter (thus > 20mm, probably 30 or whatever i'm having lying around)." ... for my 2c, the welds are possibly the critical feature (like i've posted, a 3/4" bolt is good for something like 44,000 lbs (20 tons), and if you're using a 1D thick tapped hole as a nut, that's equivalent to a standard nut.

"... They won't strip before the bolt." ... no, they'll probably strip simulataneous, depending on the grade of steel.

"Main reason is to have a sturdy backup if the castors should fail with a car on top of the bench. Max. it can "fall" is a few mm, and the bolts will then be there to avoid further failure." ... i thought the bolts were supporting the table, but this sounds like they're backing up the castors ??

if you're worried about threads stripping, why not tap the paltes only for 1/2 their thickness ... if the bolts strip, the plates will stop the table falling further, no?

still think plain holes would be better.

 

[kingnero]

Table feet is square profile tubing (80x80x5 mm), so once welded you cannot get to the "upper" nut.
Welds are not in tension, they just keep the plates in place.
Bolts will strip first, as I'm using a low-class bolt (5.8, which is the most common lowest grade). However thread stripping is not my main concern, and even if they should strip, it's only a matter of replacing the bottom plate. If the table should fall in that case, instead of only a few mm it'll fall 2-3 cm. Still not enough to throw the car of the bench. If a castor should break off, and there would be no backup, the table could fall 25 cm, enough to do some serious hart failure.

I am only backing up the castors when moving the bench from under my 2-post bridge to the side of my garage, 10 ft away.
Anytime else the bench will rest on the bolts.

I think I know what I'm gonna do now... building the thing oversize, and live with a FOS >> 10

 

[rb1957]

"building the thing oversize, and live with a FOS >> 10" ... often not a bad thing to do !

what if you had only one bolt, as a levelling means, and the other three legs sat directly on the ground; and use a car jack to light the unloaded bench to install castors to move the bench ?

maybe adding a rubber pad (to the end plates) to isolate vibrations ??

 

[flash3780]

For a wide range of steels, it's a good idea to design to 1/3 the UTS of the material when practical (or 80ksi for RC>35). Why 1/3? The fatigue limit of steels with a Rockwell C hardness of under 35 or so is between 1/2 and 1/3 of the UTS or a minimum of 82ksi(ish) for harder steels. Grade 8 bolts are generally between RC34 and RC39. Check the ASM Metals Handbook for a neat chart illustrating this.

From there, you'll want to identify all the types of loading that your joint might see. Thread shear is definitely important (Bickford's Introduction to the Design and Behavior of Bolted Joints will give guidance on calculating thread stresses). I'd wager that the workbench material will fail first (probably AISI1010 or something). The first few threads carry nearly all of the load, so if your marginal, increase the bolt diameter. I'm not aware of a significant difference in thread loading between pushing and pulling on the threads (you're looking at shearing them).

You'll also want to look at compressive stresses in the bolt (calculated at the minor diameter would be conservative).

Be sure to check for buckling (always a concern in columns in compression). Calculate the slenderness ratio (I'd use a fixed-pinned case to be conservative).

You may want to consider the case in which one of the bolt heads is not touching the floor. Your bench can be supported by only three legs.

You might want to look at the stress on the (concrete?) floor imposed by the bolt head.

You mentioned that side loads are not significant - would it be wise to review the case in which someone is moving the workbench around without setting it back on the casters?