Measuring Bolt Torque

[PeterHerman]

I am involved in an investigation of torquing procedures and we need to measure the torque of some 2" b8m bolts in a 40 hole flange. The tension was 27 kips, and the torque using the Sidmore was roughtly 600 ft-lbs. Question is: how do I measure torque? If I go around using the sequqnce, won't I affect the later bolts? Should I check in the tightening or loosening direction? Is there a digital readout wrench this large?

 

[Heckler]

Some processes and hardware that we use to check bolt torque are: using strain bolts ($$$$$$$), Preloaded Indicating Washers, Ultrasound, and Deflection Method. I appears that Ultrasound testing would work out for you.

Best Regards,

Heckler
Sr. Mechanical Engineer
SWx 2007 SP 2.0 & Pro/E 2001
Dell Precision 370
P4 3.6 GHz, 1GB RAM
XP Pro SP2.0
NVIDIA Quadro FX 1400
o
_`\(,_
(_)/ (_)

(In reference to David Beckham) "He can't kick with his left foot, he can't tackle, he can't head the ball and he doesn't score many goals. Apart from that, he'

 

[dimjim]

Probably an easy way would be to mark the bolt position
then turn it backwards one flat and then retighten it to its marked position and check the torque. This would only be a rough measurement but maybe close enough if you are averaging the readings for the 40 bolts.

I think almost no one recommends turning torques for bolts larger than 1 1/8 inches. Let us know your results.
Curious why you are doing this.

 

[rb1957]

i'd be more interested in how you know the bolt preload (unless that's just the drawing requirement). Personally, i'd be more interested in measuring the bolt preload than carefully controlling the torque. maybe look into preload indicating (PLI) washers.

agreed, tightening later bolts was an effect of the previously tightened bolts, so when you're installing the bolts you tighten they up gradually (each one a bit at a time) and then go round the joint one more time checking.

personally i don't think a digital readout would help much. if you've set the torque wrench up properly, it'll give you the torque you want.

witness marks (like dimjim is suggesting above) allow you to readily inspect the joint, to see if the bolts are lossening up.

 

[PeterHerman]

The bolts are existing, but the tightening procedure and preload are being questioned. Background: We used a 30-60-100-100 step procedure and have found that many of the bolts loosened. Some question whether the procedure was fully applied. Hence, we did a tightly "supervised" one and are going back in 3 weeks to check to see if the bolts are still tight.

I am interested in the "one flat, then retorque" suggestion...also the ultrasound (?)

On another topic, we noticed that many of the loose bolts had galling. However, the extent of galling was large. I was trying to understand, by examining the galling, what rough percentage the applied torque was lost. Any guidance?

This forum is great, by the way. Thanks for your earlier replies!

 

[rb1957]

what's the vibration loading like ?

 

[PeterHerman]

Its actually low. This is drinking water piping that is still empty prior to operation. We looked at thermal and vibration and we didn't think anything was there. We did wonder about pipe alignment and movement during installation. This is partly why we are doing the test now -- after the pipe is in place.

 

[MintJulep]

So there is a gasket being compressed?

 

[CoryPad]

In automotive, the torque on bolted joints is audited by applying a torque in the tightening direction to rotate the fastener less than or equal to 5[&degree;]. This is a quick, easy way to see if a joint has loosened. It is best to use on a high-volume process that is in control.

A better method for checking bolted joint quality is to measure the preload (instead of torque). Ultrasound (or even using a micrometer if you have access and can find one large enough) is the best method for this. You measure the initial length in the strained condition, loosen the bolt, then remeasure the length. The change in length plus the fasteners stiffness gives you the preload.

Regards,

Cory

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

 

[PeterHerman]

To mintjulep, its a hard joint with a double o-ring.

To corypad, 5 degrees in the tightening direction...this is probably what we will do, but I am still worried that it will affect the rest of the bolts.

 

[byrdj]

If it is a "hard joint", increasing the torque (thus tension) in one bolt should not effect the others. However it will if the joint is soft.

where did the 600 ft-lbs give 27k come from though

My handy dandy cheat sheet says 2" 8TPI should be about 2200 ft-lbs to get 30K

Not having all the data about oring size and durometer, but experance with flexatalic gaskets 600 ft-lbs would not compress the gasket completely and the joint would be soft.

 

[PeterHerman]

Byrdj,

27 kips per inch...on a 2" bolt I think this translates to roughly 30 percent of yield (bolt is 314 ss).

The durameter is like 70, and its 1/4" dia. But the flange has a raised face (on the non-grooved side) and the "overhand" towards the outer circumference (where the bolt holes are) flex ever so slightly.

 

[dimjim]

Are you saying that the tension for the 2 inch bolt
is 27000 pounds. I do not understand the 27000 pounds per inch. Does this mean that the 2 inch bolts has 54000 pounds tension?

As to the 30-60-100-100 step procedure, I have heard of
tightening in 1/3 and then 1/3 and then 1/3 step method
ie if the full torque was 600 ft-lbs you would tighten
first to 200 ft-lbs all around, then 400 ft-lbs the second time around and then 600 ft-lbs the last time around.

Sorry to be so dumb.

 

[byrdj]

The 27k being 30% of yeild is in agreement with my terminilogy, But 600 ft-lbs for a 2" diameter bolt is very low with my thinking

I'm not familiar with orings, so I cannot calculating the force required to compress o-ring till the flange joint goes "metal to metal". but i recall 600 to 800 ft-lbs the torque required to fully compress a flexatilice gasket so joint is metal to metal for the flanges I deal with

when I can not measure the actual strech of the bolt being tighten (something like 0.001" per active lenght of bolt gives 30K, but very material related), I use nut rotation to approxamate the desired elongation. for example if the bolts are 15" active length, and the desired elongation for 30K is 0.015", and the bolt is 8TPI, it would be rotated 1 flat. However, accounting for having the joint hard must be considered.

 

[byrdj]

dimjim

I'm use to the bolt loading being expressed as a percent of the yeild strenght of the material. thus if yeild is 90k pounds per square inch, and the % desired is 30%, then the desired loading is 27k psi. This doesn't care what size the bolt is, only what the material is.

Now if a 2" diameted bolt is loaded to 27K the actual force will be 27K over the bolt area. not accounting for threads, the area of a 2" diameter bolt is 3.14 square inches. thus expected load would be 85K

 

[dimjim]

Thanks! I understand the 87K pounds and the 27K psi.

The 30-60-100-100 is still a mystery. What does the second 100 imply?

 

[PeterHerman]

I guess that is one good question. The 27,000 lbs of tension on each bolt (~3 inch sq or 10k per inch) is sufficient to prevent flange opening during service loads. It is low, but is it too low?

However, I though the whole preload question was more complicated...like, the percentage of the force felt by the bolts during an operating event (say water hammer) was dependant on the ratio of the stiffness of the bolt & flange metals, and actually the flange absorbs a good deal of this force (so that the bolts feel, say, 60 or 70 percent of the load).

 

[byrdj]

a bolt should not "feel" any operational events. the bolt is like a spring and its initial loading should exceed any force applied to the flange. if there is any overload where the flange force exceeds the bolt clamping, the flange will open and the bolts will then feel the load.

as far as repeating 100%. when i just want to compress the gasket, that torque (calculated by gasket compress psi x gasket area per number of bolts) is used and the pattern repeated untill there is no nut rotation. then the final torque is applied


If 10K is 30% of yield, they are some mighty soft bolts

 

[byrdj]

At that low of torque, I would expect the bolts would vibrate loose very easily. are you using any locking method after torquing.

 

[dimjim]

I would like to understand your flange detail.
The O ring should be totally compressed in the groove
and you should be having metal to metal contact for
the bolts to be in tension. The Yield for these bolts
are listed at 30000 psi. Most grooves that I have designed for o rings had an area exceeding the cross
section of the o-ring. The intent is to compress the
o-ring in the groove to seal any leakage. I assume this
is a static seal design. I think the o-ring groove height
was like 80 percent of the size of the diameter of the
o-ring. I may not understand your seal design however.