Prestressing of long anchor bolts: Torque vs length? 6

[IJR]

Pals

Simple elastic strain says strain = P/EA, meaning no difference in strain will occur in a long or a short bolt as long as the same force is applied.

But some drawing spec is giving me a prestress load of 273kips(600kN) for a given anchor length and states that I should do more prestressing if I use more length

I will appreciate a guide or reference to this issue.

respects
IJR

 

[structuresguy]

Torque versus tension equations typically do not account for bolt length, as it should have no impact on the tension stress, under idealized conditions. However, the increase in length will increase the ductility (both torsionally and in tension) and the relaxation of the bolt. So I could see needing slightly higher initial tension so that after relaxation, the final tension still exceeds the required amount.

You would need to find the relaxation properties of whatever grade of steel you are dealing with, and apply some factor to increase your initial tension. I can't help you here, unfortunately.

If this is a structurally critical application, and you can not find suitable literature, I would think lab testing may be justified. The load you have listed is extremely high, and would require very large diameter bolts (3" or larger). This is beyond normal structural steel bolting requirements, so it may be hard to find design info.

But for torque values, check out this link. It has the equation to calculate bolt tension versus torque, and gives various K factors to account for different surface conditions.

http://www.efunda.com/designstandards/screws/fasteners_intro.cfm

That said, bolt tension for a given torque is extremely variable, and depends mainly on friction along the thread surfaces. Also, the equations are somewhat crude, as they don't take into account factors like thread pitch and angle.

If the pretension force is critical to successful performance, I would use a calibrated turn of nut method to ensure you are getting the pretension value you desire (I don't know if they make torque wrenches that large). And to counteract relaxation, I would consider also using a linear compression spring added to the bolt to maintain the preload force on the bolt, even after it relaxes.

 

[kelowna]

Interesting,... and difficult to know why that note has been included without more details.
Since you are talking about anchors, it might be reasonable to assume that those anchors are acting or embeeded on something different from steel (concrete, soil, rock,...) in that case I think that the increased tension for increased lenghts might try to account for possible movements on that material (which in turn would reduce the tension on the anchors). For example, in anchors on concrete, it might want to capture the long term creep under the anchor loading.

It might also be an issue with elongations. Longer bolts will elongate more for the same increase on load during service. So the designer might be worried about excessive movement (having said that, normally anchor design is such that the anchor pre-load is never exceeded so that would not be an issue).

The note might also try to account for the steel relaxation, but unless the load is very close to the anchor capacity or a very critical connection, relaxation is usually just ignored.

The designer might just be confused.

All this is, of coure speculation. If possible, the best course of action is phone the designer and ask him/her (and let us know, I am curious now)

 

[unclesyd]

Not particullary related to the specific question but might be interest.

As stated the load requires a large dia fastener of 3" or better which would normally be a single point turned bolt from preheat treated material. If one is curious here is an article that might be of interest.

http://www.superbolt.com/articles/virgin_bolt.html

Anecdotal:

Many moons before this article was written Mr John Biach inventor of the Biach Hydraulic Tensioner was at our site working on two bolting problems. During several informal meetings he was discussing bolt tensioning of very large anchor bolts used on nuclear reactors. They had tightened the bolts on a reactor and there was a question whether the proper tension had been applied on the first pass. The initial intent was to tension the bolts one time and go. When they went back with a new tension requirement they found the bolts only had about half the tension applied on the first pass. They tensioned the bolt to the new requirement and when they went back in about a week the bolt had about 90% of the tension applied. On each pass the amount of residual tension increased until it stablized at 8 passes. He stated that he had seen that problem on numerous fastener where tension was lost after the first pass. I don't think Mr Biach ever connected this relaxation to the method of fastener manufacture.

http://www.biach.com/index.html

 

[kelowna]

Very interesting Unclesyd,

Completely off topic again, but in a previous job on a previous life I had to tension and re-tension literaly hundreds of 2" dia rods (about 3 feet long). I had problems believing some the results I got. I always blamed the tensioning and checking procedure (not very clever, not my choice) and the fact that high tensile rods do not have a elastic behaviour with straight line stress/elongation relationship but a curved one with some histeresis on the tensioning cycle.

Is this the same as the 'virgin bolt phenomenom', related or completely different? I will have to think about it.

 

[dik]

Torquing can be a rough measure of the tension in the bar, but as noted, ths can be variable. Best to use a direct tension indicator or something of that ilk.

You also have to have a length in which to develop the tension where the rod is free to elongate.

Dik

 

[structuresguy]

Interesting reading, thanks for the extra info and related experiences. My own personal experience with large bolts is non-existent. But I do lots of prestressed concrete work, so am familiar with relaxation, creep, etc causing loss of prestress tension.

However, I have experienced similar problems with bolts relaxing on some longer bolts I was using on a performance engine intake plenum on my car. They were about 2 inches longer than the stock bolts, but tightened to the stock spec using a typical mechanics torque wrench. After a couple weeks, they had loosened enough that some could be turned by hand. I retighted them back to spec torque. It took doing this 4 times over the course of about a month, before they held the required torque permanently. I check them about once a month now. There could be several factors here, like relaxation, shrinkage of gaskets, etc., but the phenomenon is similar.

 

[unclesyd]

Here is a very short explanation of prestressing fasteners by different call outs, stretch, torque, turn of the nut, and stress.

http://www.riverhawk.com/boltloading.php

 

[unclesyd]

kelowna,
The initial pretensioning of a fastener, especially the longer ones, has always a point of contention among engineers and mechanics.

I imagine that if you went through a number of older papers on tightening fasteners you would see the scattering of data that could be the virgin bolt phenomena.

 

[saplanti]

I suggest not to introduce torsion to the long anchor bolts. You can use single hydraulic cylinder for this purpose or two cylinders in case you have space around the bolt to achieve this.

Today, the pretension loads are very high and torsion is always the problem side. There is always a danger to fail the bolt(s)under tension and torque. The consequence is very costly and time consuming.

However, pretensioning require anchor chair to monitor bolts in operation. If you do not use chairs you still need some free unbonded length in the foundation for pretensioning.
The bolt material should keep the ratio of Tensile Stress/Yield Stress around 1.5 and you should be in the Hook Law range to monitor the elongation under pretension. There are some bolt materials that used in the post tensioning of reinforced concrete beams with high yield strength (around 400-450 MPa). You may use them as anchor bolts as well. You need to search your local suppliers for these products to discuss further.

Hope it helps.

Ibrahim Demir

 

[IJR]

Saplanti

can you suggest literature on the subject?

respects
IJR

 

[saplanti]

DR,

You can find some information from one of the supplier below:

http://www.dywidag-systems.com/en/downloads/dsi-brochures/europe-general/

Open the file called "DYWIDAG Bar Anchors" and it has typical applications. The same web site has variety of anchor bolts and materials that you may be confused. I suggest you contact the local suppliers to discuss your needs and applications.

The chair design can be done in accordance with "AISC Steel Plate Engineering Data-Volume 2 -Part VII-Anchor Bolt Chairs". However, you may create a spreadsheet by using this reference and some structural formula from Roark's Formula for Stress and Strain.
The other two references might be Bednar's Pressure Vessel Design Handbook and Dennis Ross's Pressure Vessel Design Manual.


You may need large space above chair to fit the hydraulic cylinders in case the chair has continuous upper ring which is expected under the large loads. This sometime a bit problem if there is an access opening on the cylinder that the chair attached.

Dywidag anchor bolts do not use the usual threads and the nut lengths are a lot longer than usual for your information. The cylinder spaces might be another restraining factor in the selection of distances between anchor bolts.

Hope it helps.

Ibrahim Demir

 

[saplanti]

DR,

The other is from Canadian website with the following link:

http://www.dywidag-systems.com/en/downloads/dsi-brochures/canada.html

Click the file called "Dywidag Tie Rods". It has some material properties and the thread that I was mentioning above. I guess you can get the best information from your local supplier.

Hope it helps.

Ibrahim Demir