I am pressing a 30 inch O.D. x 2.75 3

[tc7]

I am pressing a 30 inch O.D. x 2.75 inch thick steel sleeve into a mating steel cylinder, that has an average of (-.023 inch) diameter interference. This will be a field installation and I am using 12 each equally spaced 1/2-13UNC jacking studs/nuts (120ksi U.T.S) to pull this thing together (studs go through thru-holes in a flange on the sleeve and are threaded into holes in the mating hub. Nuts on the other end of the studs are tightened to draw the sleeve in). I have calculated a total installation force required of 162,320 pounds. I have used a friction coeficent of (.18) for this press force calculation. I based this calculation on the method described in Shigley's Machine Design text.

To determine the torque required on each jacking nut, I have used 1/12 of the total press force required and used a coef of friction of (.2). Based on other equations in Shigley, I have calculated an installation torque of 223 Ft lbs to draw the sleeve into position. This is a sum of thread torque plus torque to overcome friction between the face of the nut and flange.

NOW FOR THE QUESTION - It has been well discussed that 75-80% or more of fastener torque is lost in friction and does not result in useful axial bolt loading. So I fear that the installation bolt torques that I have calculated will not really provide the sleeve installation force that I have calculated. I am suspicious that the coef of friction values that I have used may be superficially low, but I can not find justification in the literature to use a higher value, or what the higher value should be. Please advise or comment that my approach is legitimate or outline a more appropriate method.

Thanks in advance for advice or referrals.
tc7

 

[rhodie]

Man, I really have nothing to offer, other than these questions:

Are you sure your 75%-80% torque loss is figured correctly for 1/2-13UNC fasteners?

Can you press fit the sleeve in, rather than torque it down?

 

[CoryPad]

tc7,

What are the materials and coatings for all parts? Will the screws be simultaneously tightened? If not, then standard torque/force relationships will not apply.


Cory

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

 

[ietech]

Hi tc7,

I'm not sure if my suggestion will be practical for a field installation, or for the size of the components, but here goes.

First freeze the sleeve for about thirty minutes -- a bed of dry ice should do well.

Heat the cylinder as UNIFORMLY as possible just prior to sleeve insertion.

Insert the sleeve as quickly as possible --- I have had the good luck of the pieces sliding together by hand with very little effort.

I have used this method on smaller cylinders and sleeves with negative interference. It works quite well if you don't waste any time after you have started the installation.

Hope this helps,

ietech

 

[ietech]

tc7

Please disregard my suggestion above won't work on something that large. I misread your dimensions as 30" long x 2.75" dia without reading carefully, I gotta slow down a little.

ietech

 

[tc7]

Cory-
Sleeve is unplated ASTM A514, (110-130ksi UTS, 100 ksi YS)
(by the way the sleeve wall thickness is 0.65", not 2.75 as originally posted).
Cylinder is unplated ASTM A514 with wall thickness of 2.75.
Studs are unplated ASTM A193 Grade B7, (125 UTS).
Nuts are unplated ASTM A194 Grade 2H, (Rockwell C24-38).

I will use some form of lube or antisieze on sleeve and nuts.

I will not be able to tighten all nuts simultaneously. Will use a cross pattern or, just go around the circumference loading each nut up to the installation torque (or some value less than the thread shearing torque) for as long as it takes.

tc7

 

[PaulJuhnke]

Bolt stretch will help equalize the loading on the 12 fasteners (longer studs are better than shorter). Evenly increasing the torque during assembly, in a cross-star pattern, in steps of approximately 20% of maximum torque, will help prevent cocking and binding.

From practical experience, putting a good solid lubricant anti-seize on the threads and between the nut and washer will allow a nut to pull bolts in tension to failure. Without the solid lube, dry bolts will fail in torsion with an unknown tension value. Lesser lubricants create results in between the two extremes (light oil adds roughly 15% to the tension achieved).

Your interference is fairly high, which indicates you want a good grip after assembly - during operation of the equipment. The planned interference will also cause dry steel materials start to gall, which greatly increases your friction factor during assembly. Lubricants will reduce friction, helping with assembly, and reducing or eliminating galling, but they will also reduce the grip. With a reduced grip, the sleeve may walk out of position during operation. Some Lock-tite or Belzona type products will add lubrication during assembly and increase grip after they set up.

Heating the sleeve or cooling the cylinder is a common method used to assemble parts like yours, and it avoids all the problems discussed above. It’s easy to calculate the temperature difference needed. Remember to install hard stops to assist with axial location, and a method to hold the part up against the stops while cooling, because loose parts will tend to back off a bit as they cool. You also need to have insulated gloves and appropriate slings, etc.

Good Luck! Paul

 

[grinderguy]

We use .001" of press fit per inch of diameter. At this kind of ratio a 6 inch dia wheel may take 200 tons to move. If it stops anywhere but the right location and the pressing must start again, those presses such as a railroad wheel press are around 400 tons. I would be designing the press fit to be a much lower number. If rotation is a problem I would dowel it in place.
Bob Setree

Robert Setree

http://www.advanced-grinding.com

 

[tc7]

All-
I agree that the diameter interference is extreme. The fact is we have been forced into a situation where we must attempt to use an out of tolerance sleeve that has been delivered to us, rather than send back to the vendor. Therfore I need to show that the designed in jacking provision either can or cannot withstand the additional interference load before I resort to a shrinking technique or before I declare the sleeve is not salvagable.
Thanks for thinking about the problem.
tc7

 

[bradphillips]

I agree with the suggestions to properly lubricate and heat the outer cylinder as this will provide tremendous expansion on a 30 inch cylinder. Tightening 2 bolts (at 180 deg) simultaneously will help keep pieces aligned. Also check to see if the sleeve and cylinder were machined straight. ( Did you mention the length of engagement of the two pieces?).

 

[desertfox]

Hi tc7

The problem I see with your method is getting the sleeve to
go on square if the bolts you are tightening down are not done simultaneously.
Is it not possible for you to have one large screwed rod
with one nut on which passes through the center of the sleeve and connects rigidly in the mating part, you could then use a one piece round collar to press home the sleeve
and thus eliminate any out of squareness.

regards desertfox

 

[EdDanzer]

Could you use the fasteners to hold the assembly together and use a portable hydraulic jack to press the assembly together? It will help to start the assembly by heating the outside part, but speed of insertion may not be fast enough and then the parts will stick, and probably gall. Heat/cold fit is tough to do unless there is clearance for a drop in fit and alignments are good.
If you use fine thread fasteners you will have more pulling strength available. It will take about 15,000 lbs. pull to break a ½ NC grade 8 cap screw, and about 17,800 lbs, pull to break a ½ NF grade 8 cap screw.

 

[tc7]

Ed-
Is the 15,000 lb. breaking load on the UNC which you cited based on field experience? or on calculation? I can calculate ~14,800# causing yield, and ~17,000# causing rupture.

Also, I'm calculating ~150 ft lbs. torque causing the yield failure at 14,800#. Is this close to max torque values for a 1/2-13UNC in your experience?

I'm also finding that bolt tensile failure during torque-up is far more likely than thread shear failure.

I hope you can comment.
Thanks,
tc7

 

[EdDanzer]

We use the tensile stress area and the material yield strength to calculate the tension a fastener can hold when designing high pressure hydraulic cylinders. The value I gave should not fail. We have done some destructive testing to verify this method. The reason I suggested to use the bolts to hold a hydraulic jack is to eliminate torque and friction from derating the fasteners. Another advantage to using a jack is, if machining, material or surface finish variation cause the force required to press the parts together increases, just add another jack or get a larger one. We have welded fixtures to hold the jack to items that are in remote locations or large in size that need straightening or assembly. This method can be quicker, less expensive to setup and more robust than trying to bolt the jack holding assembly in place. The simplest description is a portable hydraulic press with interchangeable cylinders depending on force requirements. You may be able to rent an Enerpac (or similar type) hand or electric pumping unit and cylinder if remote pumping is required.
Good luck

 

[pmmch63]

i realize this is kinda late, but the question caught my attention.....i'm employed in a papermill and we routinely assemble parts similiar in shape and size to what you described. From my experience, seems that your amount of interference may be excessive. I don't know the length of your sleeve, but a 30" dia. only 12" long would render approx. 1131 sq. in. of surface area with .023" interference. First of all, assembling something that tight without heat will likely result in galling, without mentioning the problems that occur when that happens. Secondly, if it were even possible to assemble without heat, i would guess it would take a considerable amount of time to pull it together like you described. For what it's worth, here's what i would do....1. lower the amount of interference ( depending on exact size of cylinder) 2. try to locate an oven large enough to accomodate the cylinder (a foundry might could help) 3. heating time and temp depends upon size (again, i don't know the length of your cylinder) 4. and of course, assemble vertically if possible, allowing gravity to do most of the work. Regarding amount of interference, it usually helps to consider the forces that will be exerted on the sleeve, whether they are axial or rotational, but i think .023 is excessive in any case. If the sleeve is subject to extreme rotational force, then a key or pins between sleeve and cylinder may be needed.

 

[metman]

tc7,
better late than never?

You commented, "I'm also finding that bolt tensile failure during torque-up is far more likely than thread shear failure."
I worked with one M.E. who said they made tests that proved a nut length equal to the bolt diameter would fail the bolt in tenison rather than shear the nut. Machinery's Handbook has a very in-depth article on this subject and Tool and Mfg Engineering handbook is also helpfull. Shigley is well reputed but Machinery's Handbook also has a very good section on interference fits including both shrink and force fits and resulting retention of both types of assembly. One advantage of Machinery's Handbook is that it tends to lean toward more empiracle evaluation rather than just pure theory.

Simultaneous bolting at 180 degrees would certainly help but even if you bolt simultaneously at 120 degrees, I suspect as others have that severe galling will result with such a heavy interference. A simple test would be to use a hydraulic press with the same materials but with a nominal 2" dia using proportional wall thickness, inteference, etc. If you can actually press the pieces together without gall/welding, then you have a good chance of the real article working.

By now it is probably history so this is all academic.

tc7 are you there? What happened?

Jesus is THE life,
Leonard

 

[tc7]

Well this problem is ancient history already! Things worked out very well for us. I spent a lot of time working out a procedure to get the cylinder started nice and vertical. Before emplacing the cylinder I had coated both mating surfaces with Mobilth 460 grease, since it was reputed to have a very high pressure film rating (not sure what that means but was good to prevent galling). On the threads and contact area of the nut used to pull the sleeve in, I slopped on Locktite Moly paste which results in a coef of friction of ~.06! Additionally I placed 2 large washers between each nut and flange and all surfaces were liberally coated with the same moly paste. We torqued diametrically opposite nuts simultaneously to prevent tilting and monitored the sleeve depth on two sides all the way down. I monitored all torques and we succesfully pulled these sleeves in without ever exceeding 75 ft lbs. On some units to our surprice we did not need to exceed ~35 ft lbs torque.

We even extracted units to confirm that we could do it by the same process and that there was no galling or other damage.

It seems my torque calcs were way conservative and I attribute this to not really knowing the real friction values ( i guess you never do!) and I believe the average interferences were not all as bad as originally reported to me. It was interesting to me to see the amount of flex and conforming the cylinder experienced during the process.

Thanks for asking.

-T