ASTM A193 GR B7 bolts keep breaking 4

[mikejeffers]

Hi,
I'm new here. What a great resource.
I have a machine shop and we have been making some bull gear splice bolts for one of our customers. The material is 4142 G&P Q&T BHN 255/321 ASTM A193 GR B7.
These bolts keep breaking and I am looking for an alternate material. These studs are 1 3/16Ø by 8 1/2", turned on each end to 1 1/8" and threaded with a 12 pitch. The bolts have broken in various places but never on the bigger shank portion of the stud. Sometimes they will break at the end of the thread, but mostly they are breaking where the side of the nut meets the plates they are fastening. My customer has requested we use a material similar in tensile strength to a grade 5, even though I have told them that these bolts will break before the B7 bolts, but they insist on trying the grade 5 material, for which I am going to use 1045 material. The application sees some heat, but I'm guessing no more than 500 degrees. They started torquing the bolts at 650 in lbs, and have tryed reducing the torque to 500, but they are still breaking the bolts. There is obviously something wrong with the dryers, because they don't break on all of them. But they want to solve there problem with bolts that wont break. Any ideas on what material to try next? I would like something that is readilly available.

 

[unclesyd]

Please verify the torque value of ?650 in lbs?
What was the tempering temperature of the stud stock?
Why are you using 12 pitch threads?

Fist off you are not tightening the bolts properly if it's breaking in the first thread of the nut. You studs are apparently failing from fatigue as the bolt hasn't the preload to carry it pass it's working load.

Gr B7 is a much better choice than Gr 5 if any temperature is involved.

 

[mfgenggear]

I agree with undesyd some thing is wrong.

Is a torque wrench being used?
How are the fasterns being assembled?

are the thread Pitch Diameters or Pitch correct?
to verify correct threads are go & nogo gages being used to verify size & pitch.

These studs should be torqued upto 100 ft lbs
650 in-lbs is only 54.1 ft-lbs

see this web site nice informatiom on material & tensile properties.

http://www.rockcrawler.com/techreports/fasteners/index.asp

 

[mikejeffers]

I talked with the customer again this morning. He states that the manufacturer of the dryer sends grade 5 bolts on new machines. They suggest a torque of 680 ft lbs.
The customer says they changed from a coarse pitch to a fine pitch to gain strength at the root of the thread. (a 12TPI doesn't cut into the stud as deep as a 7TPI. We use thread wires to check the pitch diameter. They are buying grade 8 nuts from a local bolt shop like Fastenal. The temerature inside the dryer is about 500°, but where the bull gear bolts go they only see about 200°. They are sliding the stud through the hole, running a nut up one side of the stud, holding that side with a wrench, running the nut up the other side of the sted and using a torque wrench to tighten the nut to 680 ft lbs. They have tried different torque ranges, from 500 up to 680 ft lbs. Why do you say they should be torqued to 100 ft lbs, max? Maybe that's the problem right there. As for the temperature the steel was tempered at, I have no idea. We bought standard 4142 G&P Q&T BHN 255/321 ASTM A193 GR B7, from our steel supplier. We sent them some of the failed studs and they checked the material properties and said all was OK.

 

[mfgenggear]

wow 650 ftlbs , I am not qualified to give advice on stress & strain values for your bolts/studs, but that seems like a lot of torque for grade 5 bolts.


I recommend you contact a engineering consulting firm.
on the maximum torque values, depending on the material & hardness values required.

or maybe some one here can assist?

 

[unclesyd]

Just a few more questions.

What is happening to the Gr 5 OEM bolt?

Why was the nut changed?
For situation such as yours a nut of equal or slightly lower hardness would be called out? An ASTM A194 2H would do the job.

Why was the torque of 680 ft lbs chosen?
This give a bolt stress of approximately 35,000 psi which is some what low. I would use a bolt stress of 50,000 psi which is equivalent to 1148 ft lbs torque.

After the completion of the initial tightening are the bolts checked again?
On our dryers when we change a gear the bolts are initially tightened and dryer is run for a period of time and the bolts are retorqued.

Can you or your customer post pictures of the fracture surface of the failed bolts?

Again I believe if your studs are failing in first thread in the nut, indicative of fatigue, means only one thing the bolts are not tight enough for the job. They are few other mitigating causes of bolt failure but most are caused by or work in conjunction with a loose bolted connection

 

[patdaly]

What radius do you have at the root of the thread?

I assume you are single pointing this thread?

Could the break be induced by a sharp V?

 

[mikejeffers]

The root of the thread has an .008 radius. We single point the thread on a cnc lathe. The break is not where the thread ends, but further up in the threads. We have not used grade 5 in several years, but that is just what the OEM recommends. Also OEM recommends 680 ft lbs. I don't know where they got their data on grade or torque. They install the studs when the machine is still hot. The only shut the dryer down to change bolts and start back up again.
So I don't think they are retorquing after the studs equalize temperature. I tend to agree with unclesyd. They are not getting them tight enough. I think the nut was changed to a fine pitch (12tpi) in place of the coarse pitch (7 tpi) with the thinking that there is a smaller root diameter on the deeper thread, thus a smaller effective diameter, making the bolt weaker. Again, we are no longer using the grade 5, but we have gone to a 4142 G&P Q&T BHN 255/321 ASTM A193 GR B7. I'll try to get some pictures of the failed bolts and post them here. Thanks for all the help.

 

[btrueblood]

Unclesyd's value of 50000 psi tensile stress for preload is still a bit low in my opinion. A 1" to 1.5" SAE grade 5 bolt has a proof strength of 74 ksi, and torquing right to the proof load is acceptable according to many ME design texts (worst case is the bolt yields a bit, but it is doubtful that you would ever notice). For the B7 bolts, you can go to 85000 psi easily (~80% of min. yield); it's not unheard of to torque fasteners up to the min. spec. yield (105 ksi) or even beyond if it is necessary (certain auto engines have head bolts torqued into the yield region). Without the proper preload on the bolt, the threads see a much higher alternating stress component, and the fatigue life drops. In general, the higher the preload, the lower the alternating stress in the bolt, and the better the fatigue life - thus the old engineer's maxim of "torque it 'til the bolt breaks, and then back off a bit". All of this assumes that the bolt is being suitably lubed prior to assembly.

 

[unclesyd]

One other thing to check for is the dimensions of of your replacement stud. Check and see if the nut could possibly be seating/binding on the transition to the 1 3/16" cross section. This would be on the tightening end nut.

680 ft lbs for a Gr 5 would be right for a lubricated, k = .15, fastener.

There are a few other possibilities like the nut landing areas, cleanliness of the mating parts, and looking the hot bolting aspect of the assembly.

 

[mikejeffers]

I have attached a picture of the failed bolts. So, just to get this correct, They should try to torque the B7 studs to 1148 ft lbs? or even more? The picture is not very clear at all, but you can see that it is not breaking at the end of the thread, but about 1/4" away from the end of the thread, in the thread. Btrueblood, are you suggesting even a higher torque value? how much in ft lbs?

 

[CoryPad]

Your image is almost useless (no judgement, just stating a fact). It does tell us that the fracture surface is flat and occurred a few pitches from the thread runout. This points towards fatigue.

Are the bolts coated? This will affect the friction coefficient (and the required tightening torque).

Using the simple torque-tension equation:

T = KDF

where

T = torque
K = nut factor (use 0.2 if you have no better data)
D = thread diameter
F = force

For a grade 5, the proof force is 74 000 lb. A conservative preload is 75 % of proof force, or 55 500 lb.

T = 0.2 [·] 1.1875 in [·] 55 500 lb
T = 13 200 in[·]lb or 1100 ft[·]lb

It appears your tightening torque (and preload) is insufficient, which cannot withstand the applied forces, which causes fatigue cracking and failure of the bolts.

Regards,

Cory

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

 

[unclesyd]

btrueblood,
I will have to agree with you that in respect to a B7 a bolt stress of 50,000 psi is probably low for this application. I would not normally go above 65,000 for a single point thread so this is probably why I was shy in that respect. I have seen fine threads loaded to a 90,000 bolt stress at which point I holler lets go more smaller bolts.

mikejeffers,
Yes you can go up on the preload without any problems other than tightening. At the higher torque values make sure you tighten the studs is stages.

Are you replacing the studs in set or one or two at a time?

Do you have room for a torque multiplier?

See if you can get picture of the fracture surface head on with oblique lighting if possible. Looking the existing pictures the nut is coming very close to the thread runout so make sure you check the relative location of the nut being tightened. The dead end doesn't really matter but there should be a little free running thread.
As stated above it looks like fatigue even with the poor pictures.
One other thing.

Do you have a good fit on threads?

And another.

Do you have a good landing area for the nuts?
It should square to bolt axis and smooth.

 

[btrueblood]

Syd,

You may be right about sticking with a lower stress for the cut threads; but these are big bolts, and the finish ought to be inspectable. The Grade 5 fasteners I thought were plain old rolled-thread studs, not cut threads. I'd still torque the bejeezus out of them; worst case is they still fail, and your solution of more smaller bolts is the direction I would go too.

Mike,

Syd is being conservative, I'm probably being agressive (but I will routinely go through a lot of bolts, they are throw-away items to me...and I doubt you think so about the monsters you are machining).

But at least 3 or 4 of us here think the existing torque spec. is low by a factor of about 2.

 

[mikejeffers]

It costs so much more than a few bolts to have to shut the dryer down every other day than to spend a little more for the quality bolts. I truly think that the B7 is the best choice for these bolts. I am suggesting to my customer to try torqing the B7 bolts to 1148 ft bts. In the meantime, we are making them some bolts out of 1045. (gr5) The are adamant about trying this softer material. I will suggest to them that they only torque these (gr5)bolts to 680 ft. lbs. I have attached a better picture of the bolts.
Thanks everyone for your help.

 

[unclesyd]

Just remember that the torque value, 680 ft lbs, for the Gr 5 is for a lubricated fastener.
The value for the B7 of 1148 ft lbs is a dry fastener and it needs to be reduced if lubricated.

Again I would emphasize that you need to check the location of the nuts in respect to the thread runout as in you pictures it looks very close. If your nut gets into this area while tightening your torque value goes out the window.


Keep us informed as we are making something like a WAG and if we can get a little feedback it could help us in the future turning a WAG into SWAG.

 

[mikejeffers]

Yes,
I reiterated the fact about the lube to my customer. We have also lengthened the threaded length and corresponding shoulder to give a little extra room for the nut to travel.
We picked up the nuts from our local supplier, and we have found that these are also a part of the problem. About half of the 64 nuts we picked up, 36 of them are bad. The faces are not square with the thread. Of the 36, they range in face runout from .015 to .060. This is unacceptable. And could be a major contibutor to the problem.

 

[unclesyd]

Congrats, that is a major find. Something like this and little problems with the nut landing area is a recipe for disaster.
Getting the system squared up and with the proper preload a lot of your problems should go away.

I keep asking questions.
Does your gear use alignment pins at the split line?
The reason for this question is that our dryer gears and tires tend to have a lot of relative movement. Has nothing to do with our overloading same, so I've been told.

 

[mikejeffers]

to tell you the truth, i dont know if there are alignment pins or not. i have only seen the dryers a couple of times. However, knowing our customer and the conditions of the dryers, I doubt it. They have 3 out of 5 of this size dryer giving them fits. I think you're right, torquing the nuts to 1148 ft lbs, and by machining the faces of the nuts square to the thread will make a huge difference. I'm pretty sure the gear and tires don't jave a lot of lateral movement, but I have seen the tires runout in the past, but this was an old problem on a different set of dryers and was supposedly taken care of. We make them different size splice bolts all the way up to 2 3/8 diameter. Those dryers are HUGE!! You have been a great help, and I cannot thank you enough!!!!!

 

[unclesyd]

For your own enlightenment and something to keep on file is the Superbolt/Nut catalogue for your larger fasteners. Preloading is no problem with these as the smallest mechanic on the team can tightening the big boys.

http://www.superbolt.com/

PS:

I forgot to mention that your studs show attention to detail especially at the translation from thread to body.