Poisson pull out effect restraint

[UtahWater]

We are designing a 20" ID HDPE potable water pipeline. PPI's design bulletins have been useful, but seem to be somewhat incomplete in at least one area. When calculating the Poisson pullout effect force, nothing is mentioned of the restraint developed due to the soil-pipe friction along the length of the pipe. Our pipe will be buried with between 3.5 and 11 feet of compacted imported granular material.

I've not yet looked up my notes from pile-design days, which seems to be at least somewhat similar, but it seems that friction ought to be a force to be considered when calculating the required size of anchor block to be installed at a transition to bell and spigot pipe. Does anyone have any experience with pullout restraint design and the effect of friction, as I've not been able find (at least from AWWA, PPI, Eng-tips, the www) anything that mentions the restraint afforded by friction?

Also, in the second paragraph of the PPI bulletin PP 813-TN, it is stated, "...the [Poisson pullout] effect is cumulative over the entire restrained length of pipe." Yet the equation in the same bulletin for calculating the pullout force has nothing to do with length of pipe. Thus, wouldn't the appropriate word in the sentence cited above be "continuous" rather than "cumulative" or is the equation incomplete? It makes sense that length should factor into the aquation, or would a 5' length of HDPE exert the same pullout force as a 1000' length?

Any guidance on how to properly design transition restraint between continuous weld pipe and bell and spigot pipe would be appreciated!

Thank you,

 

[bimr]

A pipe section with fully restrained joints such as a long string of butt fused HDPE pipe will transmit Poisson effect pipe shortening from length to length through therestrained joints along the pipe string. Restrained joints include butt fusions,electro-fusions, socket fusions, bolted flange connections, MJ Adapter connectionsor other restrained mechanical connections. If an unrestrained bell and spigot ormechanical sleeve joint is in-line with the restrained section, the cumulative Poissoneffect shortening and possible thermal expansion / contraction effect may cause in-line unrestrained joints or connections to be pulled apart. Therefore, unrestrainedjoints or mechanical connections that are in-line with fully restrained HDPE pipemust be either restrained or otherwise protected against pullout disjoining.

For a more detailed discussion, see Chapter 4 of General Guidelines for Connecting HDPE Potable Water Pressure Pipesto DI and PVC Piping Systems TN - 36 -2006

http://plasticpipe.org/pdf/tn-36_connecting_hdpe_to_di_pvc_pipe.pdf

http://plasticpipe.org/publications/pe_handbook.html

 

[UtahWater]

Thanks, bimr.

In this particular case, I've calculated 20,000#+/- Poisson effect Force and the resulting thrust block (anchor to be poured around a molded wall anchor fitting) from our structural eng. is 10' tall, 5' wide and 2' thick with #5 reinforcement on each face. I'm OK with having this installed if it's needed, but I don't feel comfortable with it being needed. It seems that skin friction would eliminate a fair portion of the resulting force vector at the end of the pipe. The PPI manual doesn't seem to address skin friction.

Again, thanks for the response, but I still have a question on whether not, and to what degree, friction plays a role in counteracting the Poisson effect.

 

[bimr]

I think the answer is that soil friction is already taken into consideration. If you want more detailed calculations, you will need to evaluate the soil properties.

http://www.performancepipe.com/bl/performancepipe/en-us/Documents/PP814-TN Thermal Effects.PDF

Buried Piping Systems
A buried pipe is generally well restrained by soil friction along its length, and with moderate or low temperature change, soil friction alone is usually sufficient to prevent dimensional change and expansion movement. Therefore, a buried polyethylene pipe will usually experience a change in internal stress rather than dimensional change and movement. A very significant temperature decrease may exceed soil friction restraint, and apply contraction thrust loads to pipeline appurtenances. Thrust blocks for underground pipelines are usually not required unless great temperature change is anticipated.

When transitioning from DriscoPlex pipe to bell and spigot style pipes such as ductile iron or PVC, the
combination of thermal change and thrust load from internal pressure may cause sufficient contraction to pull apart the transition joint or other bell and spigot joints in the pipeline. The connection between the PE pipe and the other style pipe needs to be restrained from longitudinal pullout. Additionally, either the PE pipe needs to be restrained from longitudinal movement (in-line anchor) or a sufficient number of upstream (or downstream) bell and spigot joints need to be restrained against pull out. The manufacturers of ductile iron and PVC pipe typically rovide methods for calculating the number of joints that need to be restrained for a given axial force.

If temperature change is extreme, low thrust capacity (unrestrained) connections to manholes may require longitudinal force thrust block (in-line anchor) protection. See Figure 6.

The longitudinal stress from temperature change may be estimated using Equation 2. Soil load bearing capacity will require appropriate soils testing. Temperature changes below grade usually are not instantaneous, so an appropriate long-term elastic modulus from Table 1 should be selected. Figure 6 illustrates a typical thrust block design.

 

[rconner]

See also prior discussion at

http://eng-tips.com/viewthread.cfm?qid=183726.

In various applications there are as many as three separate and potentially significant drivers to contracting movements of hdpe pipelines i.e. “recoil” from any high pulling load HDD operations that might be involved, thermal contraction when the pipe temperature becomes lowered from assembly condition, and also Poisson-effect contraction when the pipeline becomes pressurized. While it can certainly be argued soil moderates at least some of these (in some cases cumulative) movements or loads, anchor collar(s)/flange(s) directly on the pipe that’s generating same, and a good-sized thrust wall around same for anchorage, does not sound like a bad nor unfair idea to me.

 

[UtahWater]

Thanks, bimr and rconner, for the input.

 

[UtahWater]

FYI - I just spoke to a 30-yr veteran engineer in HDPE and he says that he's never seen pullout due to Poisson effect in a buried application. Pullout due to not letting the pipe come to thermal equilibrium, or not letting it relax after a pull, yes, but no on Poisson's.

 

[bimr]

That person's opinion may be true.

However, the recommendations put forth by the various manufacturers are most likely based on actual material properties and are probably more reliable in the long run. Facts are stubborn things.

What the person is telling you is equivalent to hearing how one's automobile seems to run better after it is washed.

 

[rconner]

This is indeed an interesting statement from “a 30-yr veteran engineer in HDPE”. I am aware however that this subject was covered to some extent in a well attended 2010 Thrust Restraint workshop at the annual conference of the Pipeline Division of ASCE in (it appears your neighboring) Colorado, with the following verbiage included in a draft White Paper that I, along with all other attendees, received at this event:

“Special considerations are required where PE pipe is joined to other pipe materials that are in turn joined with unrestrained joints. In such cases the pipe will shorten upon pressurization. The shortening may be sufficient to pull apart unrestrained joints in connecting pipelines. To prevent this, the PE pipe must either be securely anchored (as with an adequate fused collar embedded in a reinforced concrete thrust wall) to withstand all forces and movements generated in the PE pipe section, or a sufficient number of joints in the connecting pipeline must be restrained to prevent movement.”

I believe at least one of the gentleman on the Committee that reviewed this subject and this behavior has at least that amount of experience (if not more), and with association incidentally of/to some very well-known manufacturers and applications of hdpe pipes. For whatever it is worth, I have myself watched the pressure testing of quite a few different types of plastic pipes over a great many years (though admittedly in my case in a laboratory, as opposed to underground setting). Every pressure test I ever witnessed with unrestrained (e.g. gasketed slip-on or unrestrained mechanical joint) closures appeared to exhibit PRONOUNCED axial movement or shrinkage at such closures as pressure increased. While I would not purport to have the expertise of either of these gentlemen, with my admittedly perhaps some limited grasp of the mechanics of (long term) very low modulus and (long term) very high Poisson ratio materials, I attributed what I saw then with my own eyes (when some pipe ends even pulled inward past the seals etc.) to “Poisson effect”.

Everyone have a good weekend.

 

[UtahWater]

rconner, do you happen to have links to the presentation and/or white paper from that conference? I haven't yet had a call back from Performance Pipe engineering to get their take on this question that I haven't seen addressed in any documentation. I would be interested in seeing what research has been done regarding the effect of skin friction on the Poisson effect. Thank you!

 

[bimr]

Performance pipe does address Poisson Effects in the publication at this line:

http://www.performancepipe.com/bl/p... Mechanical Restraint and Poisson Effects.PDF

"In pressure water systems where DriscoPlex™ HDPE pipe is connected to non-HDPE components such as belled PVC or ductile iron pipe, belled fittings and components, and mechanical joint bell pipe, fittings and components, the connection must be restrained to prevent Poisson force joint pullout."

"When pipes made from ductile materials are pressurized, the diameter expands slightly and the length decreases in accordance with the Poisson ratio of the material. With unrestrained bell and spigot joined lengths, the effect is limited to the individual pipe lengths, but with fully restrained pipes such as fusion-joined PE pipe, the effect is cumulative over the entire restrained length of pipe. When fusion-joined PE pipe is connected to unrestrained mechanical couplings or bell and spigot joint PVC or ductile iron piping, Poisson effect pipe shortening can cause pullout disjoining of unrestrained joints where the PE pipe transitions to the unrestrained non-PE pipe. To prevent Poisson effect pullout disjoining in the transition area, provide protection by installing external joint restraints at unrestrained bell and spigot joints, or by installing an in-line anchor in the HDPE pipeline, or by a combination of both techniques."

 

[UtahWater]

Thanks, bimr. Yes, I've studied PP813-TN lately. I'm not suggesting that the Poisson effect doesn't apply to buried pipelines, I'm looking for a specific mention of how skin friction effects (affects? I'm never quite sure) the pullout results of the Poisson effect. In the case of a buried pipe in a compacted trench, is the result of the Poisson effect an increase in the stress in the pipe wall, rather than movement (or force, if restrained) at the pipe end? Some combination of increased stress and reduced movement (force)?

I'm also lacking insight into why, on page 3 of PP813-TN, the equation does not consider length of pipe, since "the effect is cumulative over the entire restrained length of pipe."

Is bell and spigot pipe not really restrained, since it is recommended that only some of the joints be "restrained"? If they were really restrained, wouldn't the Poisson effect carry on through the restraints and, in effect, not stop until there was an anchor block to resist the movement...or skin friction on the pipe, maybe?

Either I'm really dense, or really ignorant. Either way, thank you all for your input and helping me understand.

 

[bimr]

Bell and spigot pipe is not a restrained pipe unless you have the thrust lock restrained joint. As long as you have bell and spigot installed in a straight line, you do not have to worry about thrust.
When pipes made from ductile materials are pressurized, the diameter expands slightly and the length decreases in accordance with the Poisson ratio of the material. With unrestrained bell and spigot joined lengths, the effect is limited to the individual pipe lengths, but with fully restrained pipes such as fusion-joined PE pipe, the effect is cumulative over the entire restrained length of pipe.

See this ASTM document:

http://www.astm.org/Standards/F2164.htm

Poisson Effect—When test pressure is applied to plastic piping systems that have fully restrained joints (joints such as heat fusion, electrofusion, bolted flanges, and so forth.), diametrical expansion of the pipe may reduce the overall length of the fully restrained section. Poisson-effect length reduction may affect or cause disjoining in other contiguous sections that have partially restrained or non-restrained joints, such as bell-and-spigot joints, when such joints are in-line with the test section. To prevent Poisson-effect disjoining, take measures such as the installation of external joint restraints (diametrical clamps and tie-rods) on in-line non-restrained joints, installing in-line thrust anchors at the ends of the fully restrained section, or isolating the fully restrained test section from piping with non-restrained or partially restrained joints.

http://www.pfdistributors.com/Municipal-Industrial/mechanical-restraint-poisson-effects.html

 

[rconner]

I don't believe there is presently a link to this White Paper; also, I don't recall the polyethylene folks favoring us with specific test data regarding soils and skin friction along their pipes (though one would think at some point over many decades someone would have looked at this).
However, while they are not necessarily the same animal steel water pipe folks in AWWA Manual M11 talk about coefficients of friction in the range of 0.25-0.4 for this buried pipe, andthis pipe is of course often "polyethylene" (adhesive)tape-wrapped (and I wouldn't doubt at least the lower range could be referring to such in some soils), and DIPRA and EBAA Iron have of course tested polyethylene-encased ductile iron pipes that are even more widely used in various soils, and I believe their recommendations and other information for different soils are freely available on the web.