Hi
This is my first post on this Forum. I'm interested to see what responses there are to my query as I'm having difficulty getting an answer from Dr Google.
My problem is as follows:
We have been asked to certify a clamp. The clamp is a steel collar around a mm DIA reinforced concrete pier, with a four (4) bolt configuration each side of the clamp. Each of the four bolts are proposed to be tensioned to 200 kN. The collar is designed to resist vertical sliding on the pier.
My question is this: What is the sliding resistance of the clamp, assuming the coefficient of friction is 0.3? Is it 200 x 0.3, or is it 4 x 200 x 0.3?
I would have said that the sliding resistance is 200 x 0.3, as there is only 200 kN of total tension in the system (despite there being four bolts). However, other engineers have claimed that we should allow 4 x 200 x 0.3 (i.e. the bolt clamping forces are additive).
It would be great to hear people's opinions on how to design such a connection and also if there are any useful references available for pipe clamp design. As I mentioned, I am really struggling to find any reference material. I'm happy to provide more information if any of the above is unclear.
Thanks!
Riser clamp design, especially for something as large as this ( 2.4m diam pier with bolts with a tensile load of 20 tonnes is a specialist design activity and often needs FEA. The chance of the two diameters (pipe and clamp) being exact is next to zero and hence you then have an intermediate material to even out the bumps.
What is the shape of this clamp? How much room is there between the two halves?
It's not a easy as F x friction.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
Hi LittleInch
The clamps are semi-circular in two halves, with a 250mm gap between the halves on each side of the configuration. The four bolts are each side of the clamp configuration. When the bolts are stressed, the two halves are pulled together to close the gap. Hope that makes sense.
If the first bolt is stressed to 200kN, when you stress the next bolt to 200kN you are not necessarily putting more load into the system. You are just bringing that bolt to the same load that is already in the system. Or is that wrong? It would be great to get a reference for designing these types of connections.
The bolts are proposed to be stressed in stages - 50%, 95%, 100% plus a final round of tightening / checking to ensure the bolts have the same load.
Should we recommend that the configuration is load tested if it is not a simple F x friction calculation?
Thanks again!
swengi said:
If the first bolt is stressed to 200kN, when you stress the next bolt to 200kN you are not necessarily putting more load into the system. You are just bringing that bolt to the same load that is already in the system. Or is that wrong? It would be great to get a reference for designing these types of connections.
Imagine you have a cable hanging vertically. Add a 20,400 kg load to the end of this cable. The cable tension is now 200kN.
Add another 20,400 kg mass to the end of your cable.
Is the cable tension still 200kN?
Nope.
Your scenario is, however, much more complicated than that. When you tighten your first bolt, you stretch the clamp and bring it into very firm contact with the pier, over some or all (depending on how stiff the clamp elements are and how closely they match the shape of the pier) of the diameter of the pier surface.
When you tighten the second bolt, you add more tension- but this tension is trying to stretch the clamp against the tensile stiffness of the clamp itself PLUS whatever additional stiffness is added by loading the surface of the pier in shear, all the way around.
The normal force between the clamp and the pier is going to be somewhere north of 200kN and somewhere south of 800kN. Finding out what it actually is (and where, because it won't be the same everywhere in the real world) is very difficult without FEA.
Hi jgKRI
Thanks for your response. This is where I was trying to get to. I understand that the answer will not be simply 200kN or 800kN but rather will lie somewhere in between.
I was hoping there might be some method of analysing the connection other than FEA, although it sounds like it may be too difficult to model without FEA.
In the past I have used a coupler (clamp) system for scaffolding design to attach scaffolding tubes. Apparently adding an additional 'check' coupler clamped beside the first coupler would increase the sliding resistance by a factor of 1.5, not 2. However, this was just an empirical factor quoted by a principal engineer. I haven't been able to find any published material on how these factors are calculated.
The only other way to calculate the capacity would be to load test it. However, this will likely be too difficult and the results may not be meaningful for a sample size of 1.
Thanks!
How much load is applied to this, and is it a life safety application? The answer regarding what to do here is very different depending on those answers.
On an untested, theoretical application, I wouldn't be using a 30% friction factor, or I'd be using a high factor of safety on top of it. I'd also be conservative to account for thermal effects that might occur based on the different materials and possibly long term corrosion of the steel or something like local surface wear on the concrete if this thing is there a long time. I would also want there to be redundancy in the structural system such that one failure of a clamp doesn't result in a catastrophic failure.
Unless you're in a short term application with periodic inspection, I wouldn't be using scaffolding design as a reference for factors of safety or exactly how to deal with this, although it's obviously a nice reference to be aware of for general design practices.
If you haven't seen it already you could try this.
It's based on using a neoprene liner which is common for steel to steel riser clamps offshore, but might give you a starter.
All offshore clamps of a decent size end up with some sort of FEA normally.
Four bolts on a clamp are going to bend and stress the clamp in certain ways and if the total stretch on the clamp is that much there will be friction effects going on and once you close up the gap then the bolt load isn't going into the clamp to pile contact anymore. You also have temperature expansion issues on the metal clamp.
Can't you just drill a hole through the concrete pier??
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
Hi swengi
Yes the principal engineer is throwing out a ball park estimate of would could be obtainable, to throw another complication in you should really tighten the bolts is a sequence because tightening one and then an adjacent bolt could actually reduce the tension in the first bolt that was tightened, it comes down to how accurate you need to be.
&#;Do not worry about your problems with mathematics, I assure you mine are far greater.&#; Albert Einstein
swengi said:
I was hoping there might be some method of analyzing the connection other than FEA...
There is. Because of the assumptions that have to be made (as noted by other members), the answer is useful only as an upper bound of the normal force. This upper bound is surprisingly straightforward: F = 2 pi T.
Multiply the normal force by the coefficient of friction to get the upper bound of the friction force.
See: "Force Exerted By A Band Clamp"
The author provides both a simplified and a rigorous mathematical analysis that both result in the above equation.
As a 's bridge contractor we uses friction clamps as described by the OP (but smaller, 36" diameter concrete columns) to support formwork for bridge pier caps under construction. Today, safer concrete inserts are for cap construction. This is a picture from the web shows the concept:
SRE - The crucial issue for me though is whether a tightening band is truly the same as two half clamps being clamped together? The force works on the basis of a uniform change in circumference.
A clamp in two halves works on the basis of force applied over the contact area.
Similar but truly the same thing?
And do you use the tension of two bolts or four?
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
"And do you use the tension of two bolts or four?"
However many there are at each clamp connection. If I understood the OP correctly, for the clamp under consideration there are 4 bolts on each side of the shaft, bolting 2 halves of the clamp. In that case, you would count the tension in 4 bolts, since that is the magnitude of the tension at any and every point around the clamp (neglecting friction, of course).
Thank you,
HotRod,
dhengr &
bridgebuster.
LittleInch - I thought about clamp fit before posting. For our bridge construction work the fit was very good:
1) We had permanently employed, experienced (bridge) concrete workers.
2) The forms were steel, designed and manufactured specifically for 36" diameter, high-head, concrete column construction (20' high, PSF fluid concrete pressure).
3) The clamps were designed and manufactured by the same company that made the steel forms: "Economy Forms" now known as "EFCO".
4) The two halves of the clamp are each a few degrees short of a 180o arc. When fully tightened, there is an intentional gap between the two halves. By checking tension on all bolts, on both sides of the clamp, tension in the clamp can be equalized fairly well.
5) The image shown below is strictly conceptual. The clamps look and perform like the professionally designed structural members that they are - not a simple "band clamp".
.
As I understand the
OP's question, there are four bolts, in parallel, on each side of the clamp (eight bolts total). Each of the eight bolts is tensioned to 200 kN - all bolts tightened in small increments -
not "one at a time". Total tension in the fully assembled clamp will be essentially 800 kN.
Even so, I would expect a Safety Factor in the 4 to 6 range to be applied. SF of 4 to 6 is common in other aspects of heavy construction.
Hi SlideRuleEra
You are on the money. That is the exact application this clamp is being used for (short-term). We are supporting the pier headstock formwork off the clamp, in addition to a couple of levels of working decks for access.
I did use the paper you cited in our design checks and the F = 2 pi T formula. The friction factor I used is 0.3, which comes from the Australia bridge code AS . The friction factor they quote is 0.6 for steel on concrete and they use a 'phi' factor of 0.5 (i.e. FOS of 2). So 0.6 * 0.5 = 0.3. I understand that friction in this application is not quite as straightforward, hence my question regarding any references for designing these specific types of connections.
We have an additional FOS of 2+ on the connection. So assuming the FOS on friction is correct, the overall FOS would be in excess of 4.
Thanks for the helpful response!
Is there such a thing as a pipe clamp that will allow the pipe to run parallel with a piece of strut instead of right angles as below? I want to call it a twisted pipe clamp.
There&#;s probably an obvious solution to this issue, but I can&#;t think of it this evening.
Thanks
You mean like a parallel pipe clamp? We use these with EMT.
Parallel Pipe Clamps are designed for use with common conduit, pipe and tube sizes when needing to run parallel to the channel. For application examples, refer to our Application Showcase.