PVC Design Values

[sda97]

Are there any published stress values for pvc piping? I have found some values. Specifically, I am looking for shear (not torsional) values and yield strength. I am lookng to see if is possible for the a pvc pipe to shear off through the cross section of the pipe.

 

[PrentP]

sda97

Have a look at ASTM Standard D1785-03 entitled "Standard Specification for Poly(Vinyl Chloride) (PVC) Plastic Pipe, Schedules 40, 80, and 120". There are additional references in this standard.



Regards,

Peter P

 

[sda97]

Thanks for the information. I did find some values on som manfacturer's websitesn

 

[EdClymer]

Sda97


Here in the UK, we use BS EN1778 to establish a safe stress - with raw data derived from 40 years of long term testing of pipe.

Would need service temperature and life expected.

Regards


Ed





Ed Clymer
Resinfab & Associates
England

 

[sda97]

The two estimated service temperatures for the pvc pipe are at issue. One is 0 degrees F and 60 degrees F. Thank you.

 

[RichGeoffroy]

Sda97:

The maximum recommended design stress for rigid PVC (PVC-u) such as that used in PVC pipe (ASTM D1785) is 2000 psi. At this stress, a properly manufactured PVC-u part should not fail --- at least not in our lifetime. The material can be expected to withstand a stress which is twice the design stress, 4000 psi, continuously for about ten years prior to rupturing.

Is it possible for PVC pipe to shear off through the cross section of the pipe? Yes, but not under any “normal” circumstances. In the instance where there is a shift in the soil, such as might be caused by collapse of a sink hole or an earthquake, the pipe must resist much of the weight of the overburden soil. In such situations, PVC pipe can fail under shear. PVC pipe can also be sheared through its cross section by a backhoe during excavation. In these cases the tensile yield stress of PVC-u, 7500-8000 psi, can be used to calculate the load necessary to result in such deformation. However, these are not considered to be normal situations.



Rich Geoffroy
Polymer Services Group
POLYSERV@aol.com

 

[McGuyver]

Hello Mr. Geoffroy,

first off, thank you for your very informative posts(in general), they are very helpful. Sorry to derail this topic but I have a dilema I hope you might have some insight on, which is relate to your previous post.

I've read through ASTM Standard D1785-03 per your suggestion and also read through the PPI publication TR - 4 - 03 to get a feel for design guides for PVC. I noticed that these all are written for and tested for PVC pipe and fitting designs where the major stresses seen are tensile stresses in the circumferential or hoop direction. I am working on a project that involves a design that incorporates basically a flat PVC plate with some upward curvature and reinforcing ribs top and bottom. This is subjected to relatively low pressures however acting over a significant amount of area on the bottom side. Anyway, this results in a more complicated stress state involving bending stresses than your simple PVC pipe problem, and I was wondering how much these design guides for PVC pipe still apply in cases like this?

I need the PVC "plate" area to remain fairly stiff (to minimize deflections) but as I have seen, most PVC materials used in pipe construction are fairy ductile lending to large deformations. The only option I see is to use a glass filled PVC material but this has some drawbacks from a processing and cost standpoint. However, our material specialists pointed out a PVC grade from the PPI TR - 4 - 03 that is rated as having been tested for use in pressurized pipe and fittings applications. He says that they don't have any "impact modifyer" added which gives them a lower notched izod value, but makes them "stiffer than what the flex modulus value on the material data sheet reflects." I'm a little leary of trusting to use this material based on some unquantified asertion as to its alleged stiffness. Especially since the data sheet value for flex modulus says otherwise and our case requires higher bending stiffness from the material than typical unfilled PVC. Well, any insights, helpful input, or recommendations will be greatly appreciated.

 

[RichGeoffroy]

McGuyver:

Hoop stress is just another tensile stress and it is tensile stress which causes materials to fail. You can use the long-term creep-rupture data developed on pipe for any other product --- so long as you can accurately calculate the actual tensile stresses in your part. Remember, however, that the long-term design stress is compound specific. Meaning that alteration of the compound, even using the same base resin, may significantly alter the long-term performance for that compound. So, unless you are actually using a stress-rated PVC pipe compound, you should consider using a higher safety factor, or a lower design stress.

As for impact modifier, Type I rigid PVC which is typically used in conventional PVC 1120 pipe should not contain any plasticizers or impact modifier, however, most contain processing aids which may be similar to impact modifiers. Type II PVC compounds are impact modified and exhibit higher impact, lower tensile strength, and lower modulus than the Type I PVCs. See ASTM D1784 for the various general properties for the different grades.

When it comes to reinforcements, so long as the glass has a suitable surface treatment and is well dispersed in the resin, long-term failure will result from failure of the matrix. The stress for long-term part failure will be higher than the virgin compound because of the reinforcing effect of glass fibers, however, the strain-to-failure in the matrix will remain the same. For long-term applications, keep your maximum initial strain between 0.5 and 1%. You’ll note that most of your design stresses fall in this strain range.

Finally, don’t trust data sheets. Use them as guides in helping you make your decision on potential candidates, but do your own tests. Collect and analyze the data yourself --- don’t just get the “numbers” from the lab. Make some sense out of the behavior of the different compounds and make a sound engineering decision as to the best candidate for your application.




Rich Geoffroy
Polymer Services Group
POLYSERV@aol.com

 

[McGuyver]

You can use the long-term creep-rupture data developed on pipe for any other product --- so long as you can accurately calculate the actual tensile stresses in your part.

I'm assuming that this means calculating tensile stresses by way of a Finite Element Analysis (given a set of loads, restraints, & material properties.) This brings to mind another consideration: As I understand it, plastics exhibit pretty non-linear behavior not only in their stress/strain curve but they exhibit different behavior under different loading conditions, which is why I'm assuming the separate flex modulus is given. If my primary stress on the part is due to bending, then shouldn't I use the flex modulus? I've heard some people say that they'd rather use the tensile modulus even in bending because the tensile modulus is usually a little bit lower than the flex modulus and by using it, it lends a little more of a safety factor in their design.

For long-term applications, keep your maximum initial strain between 0.5 and 1%. You’ll note that most of your design stresses fall in this strain range.

Are you talking about calculated strains from an FEA, or measured strains under actual loading using a strain gauge?

Thank you again for your knowlege and valuable insights on this matter.

 

[RichGeoffroy]

Failure is the result of tensile stresses (see Plastics Engineering FAQ, “How do Plastics Parts Fail?”). Essentially a material’s modulus should be the same whether it’s measured in tension, flexure, or compression. The problem arises from the way the tests are conducted and that creates the variation. Tensile is probably the purest form of measurement, so I would rather use tensile data (see Polymer Engineering FAQ, “Is Flexural Modulus different from Tensile Modulus and Compressive Modulus?”).

Certainly, good FEA data will provide the best strain information, but it should be checked with strain gauge data to ensure the quality of your assumptions. Strain gauge data is good data, but it has its limitations. First, you can only get strain gauge data in an area where you can apply the gauge. Secondly, strain gauges average the actual strain over the area of the strain gauge, so the data may be somewhat misleading. Finally, strain gauges tell you nothing about the high strains that occur in localized areas, such as sharp corners and other stress concentrations where failure will most likely initiate.

Keep in mind that we were talking about “maximum initial strain” between 0.5 and 1.0%. This initial strain defines the stress for the constant stress that the part will experience over its lifetime. In actuality, the strain at failure will be considerably larger due to creep. Under a condition of constant strain, the part can experience a somewhat higher initial strain because the stress it creates will relax with time; but this initial design strain is bit more difficult to estimate. The general rule-of-thumb is to not exceed 2% strain.




Rich Geoffroy
Polymer Services Group
POLYSERV@aol.com

 

[McGuyver]

Thank you Mr. Geoffroy,

You posts have been very helpful. I've read through several of your FAQ, and they've been very illuminating on these subjects. I appreciate your contribution to this forum.