We are estimating lateral deflections for a socket in yielding rock and have developed p-y curves to take into account the potential for lateral shifting at the top of rock. I am uncertain as to what to enter for the soil modulus parameter k (pci) because I do not know how the program uses the factor. According to the program documentation at Chapter 4, k is the slope of the modulus E versus depth. This may be useful for soil, but the modulus is relatively insensitive to depth for rock. Examples included with the program show higher k factors with increased soil material stiffness, e.g., 90 for medium dense sand but 2000 for weak limestone. This would suggest a greater increase in mudulus with depth, not a lesser increase. I have access to relevant subgrade modulus values tabulated from plate loading tests, etc. for rock, but am hesitant to specify a k-factor that may mean something in the program that does not mimic reality. Does anyone know how to get COM624 to resolve this?
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
Modulus K-factor in COM624p for rock
- Thread starter sgsibob
- Start date
Does anyone know how to get COM624 to resolve this?
Yup. Well, sort of.
COM624 uses the k-value to limit the initial slope of the p-y curve. If the initial slope of the p-y curve is too steep, you can get divide by zero errors. Theoretically. Practically speaking, it isn't an issue.
You can handle the problem a number of ways:[ol][li]Divide the rock into lots of layers, and vary the k-values so the k*x = E[sub]rock[/sub] (x is depth below ground surface.)[li]Input your own p-y curves - this bypasses the k-value issue completely.[li]Pick a reasonable k-value for the top of the rock, and then ignore the k-value unless you have a problem.[/ol]
I hope this helps -
![[pacman]](/data/assets/smilies/pacman.gif "[pacman] [pacman]")
Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.
Yup. Well, sort of.
COM624 uses the k-value to limit the initial slope of the p-y curve. If the initial slope of the p-y curve is too steep, you can get divide by zero errors. Theoretically. Practically speaking, it isn't an issue.
You can handle the problem a number of ways:[ol][li]Divide the rock into lots of layers, and vary the k-values so the k*x = E[sub]rock[/sub] (x is depth below ground surface.)[li]Input your own p-y curves - this bypasses the k-value issue completely.[li]Pick a reasonable k-value for the top of the rock, and then ignore the k-value unless you have a problem.[/ol]
I hope this helps -
Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.
- Thread starter
- #3
I developed a p-y curve but there is nothing in the program docs that says you don't have to have a K value to start with. I would like to dispense with K altogether -- you mean the program won't blow up if there is no K value entered? (Example 5 with the documentation shows user input p-y values and also has a k value entered, which is why I'm confused).
We need a better rock socket program than Rocket -- which the developers acknowledge is proven only for soft rock sockets in Melbourne, I think. That's why I am suffering with COM624p.
We need a better rock socket program than Rocket -- which the developers acknowledge is proven only for soft rock sockets in Melbourne, I think. That's why I am suffering with COM624p.
If the program insists on a k-value with hand crafted p-y curves, then give it a very large value so that it doesn't interfere with your manually entered ones.
I like COM624 - did my master's research on laterally loaded pile groups. But it isn't very 'user friendly', and can be a real pain sometimes. It does, however, give good results and has a well established track record.
How close is the rock to the ground surface?
Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.
I like COM624 - did my master's research on laterally loaded pile groups. But it isn't very 'user friendly', and can be a real pain sometimes. It does, however, give good results and has a well established track record.
How close is the rock to the ground surface?
Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.
- Thread starter
- #5
The rock is buried below about 15 ft of sand/gravel and cobble terrace deposits.
The thing is that the bedrock upper surface was exposed to surface erosion that created steep to vertical fissures several inches to over a foot in width, that became filled with fine air-blown sand before deposition of the overlying terrace deposits. We are trying to account for this because it holds the potential for lateral translation of the rock blocks along basal discontinuities that were detected in the borings and observed at other locations in the field. The influence of fissures depends on the dimensions of the blocks with relation to the micropiles. The fissures are irregular (they are not karst, by the way) but narrow to weathered joints with depth. At places it can be over 25 ft below the rock surface before the fissures pinch out entirely, but this is uncommon. We are modeling the blocks as 5 ft high and 4 ft across, separated by fine sand filled fissures 3 in thick. The fine sand is modeled as medium dense but stiffening as the block closes the fissure horizontally. The P-Y curve reflects displacements under the sum of the elastic block deformation and the compression of the sand fissure resulting from the residual lateral force left over after deducting for basal shear.
This is where the k-factor comes in. Initially, the k-factor would be the subgrade modulus of the intact rock because the small strain deformations are entirely due to the elastic deformation of the block, with no inelastic shifting. The P-Y curve accounts for the inelastic portion. It's a simple enough matter to calculate the initial subgrade modulus and it is indeed high, like 10e03 pci.
Think that will work?
The thing is that the bedrock upper surface was exposed to surface erosion that created steep to vertical fissures several inches to over a foot in width, that became filled with fine air-blown sand before deposition of the overlying terrace deposits. We are trying to account for this because it holds the potential for lateral translation of the rock blocks along basal discontinuities that were detected in the borings and observed at other locations in the field. The influence of fissures depends on the dimensions of the blocks with relation to the micropiles. The fissures are irregular (they are not karst, by the way) but narrow to weathered joints with depth. At places it can be over 25 ft below the rock surface before the fissures pinch out entirely, but this is uncommon. We are modeling the blocks as 5 ft high and 4 ft across, separated by fine sand filled fissures 3 in thick. The fine sand is modeled as medium dense but stiffening as the block closes the fissure horizontally. The P-Y curve reflects displacements under the sum of the elastic block deformation and the compression of the sand fissure resulting from the residual lateral force left over after deducting for basal shear.
This is where the k-factor comes in. Initially, the k-factor would be the subgrade modulus of the intact rock because the small strain deformations are entirely due to the elastic deformation of the block, with no inelastic shifting. The P-Y curve accounts for the inelastic portion. It's a simple enough matter to calculate the initial subgrade modulus and it is indeed high, like 10e03 pci.
Think that will work?
It sounds like you have a good handle on the problem - it should work just fine. If it doesn't, post another message with the details of your problem - and I'll try to help.
There is a back-up approach - use a program called BEAMCL28 that Mike O'Neill produced. It is much "cruder" than COM624 - you have to give it all the information manually, no automatic parameter generation. But since it is a more general routine, it does a better job of handling the odd problems. I like to use it for Q-W analyses of marine bulkheads.
Look for it on the internet, it may be in the public domain. Also check with FHWA - if you are interested or need it.
Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.
There is a back-up approach - use a program called BEAMCL28 that Mike O'Neill produced. It is much "cruder" than COM624 - you have to give it all the information manually, no automatic parameter generation. But since it is a more general routine, it does a better job of handling the odd problems. I like to use it for Q-W analyses of marine bulkheads.
Look for it on the internet, it may be in the public domain. Also check with FHWA - if you are interested or need it.
Please see FAQ731-376 by [blue]VPL[/blue] for tips on how to make the best use of Eng-Tips Fora.
- Thread starter
- #8
- Status
Similar threads
