It is regrettable that you are having field problems with at least the thinner pvc pipe, and it would appear relatively early in the intended life of such infrastructure. I am not going to attempt to prove exactly what has caused the problems you are experiencing (that is perhaps the realm of high paid forensic engineers!) However, I will briefly define some issues that may help in understanding. Your quite very short question is perhaps more complicated than it sounds, particularly if one wishes to visit all the issues potentially involved. When you say “SDR 35” and “SDR 26” pvc pipe, I believe you are not necessarily just talking about two different thicknesses of pvc pipe, but also pipe potentially manufactured to somewhat different standards with at least slightly differnt expectations (I’ll get back to this later).
DR or SDR is of course however basically the ratio of the outside diameter of the pipe to the wall thickness. E.g. it follows that an 8” SDR 35 pvc sewer pipe per say ASTM D3034 with an exterior diameter of 8.40 in. would have a wall thickness of 8.4/35=0.24 in. It would also follow that an 8” SDR 26 pvc pipe, per say ASTM D2241, with a slightly different exterior diameter of 8.625 in. would have a wall thickness of 8.625/26=0.332 in.
The conventional pipe stiffness (or ring deflecting/crushing strength) of a pipe is as determined in 3-edge or parallel plate laboratory testing, basically the load vs vertical deflection from that load per inch of ring, or it can alternatively be conventionally calculated or approximated from the material modulus, wall thickness, and diameter properties of the specific pipe/ring involved (using e.g. the ring formulae as also depicted in e.g. ASTM D2412, for external loading/parallel plate testing of plastic pipes etc.)
Now, it must be explained with regard to the stated question that with the latter/mentioned ring formula the relative pipe stiffness is not directly proportional to just the DR or the thickness, but instead the CUBE of the thicknesses (as it is the rectangular moment of inertia of the wall cross-section that is involved in the, in effect curved beam, determination). If the material and geometries of the pipes were identical, in theory therefore the SDR 35 pipe in question would have only 0.24^3/0.332^3 (100%) or <38% of the stiffness of the SDR 26 pvc pipe.
It is however quite possible that the specific pvc materials/SDR’s you ask about (due to different ASTM standards etc.) may not be exactly identical, and that could throw these values off just a little. In this regard I believe some experts have claimed some manufacturers in the past have had a tendency to “load up” sewer pipes pretty heavy with a high percentage of cheapening fillers like calcium carbonate (~limestone?), due to the fact that this requires less resin/makes the pipe cheaper to produce/higher profits and doesn’t necessarily hurt the ring strength of pipes either (it might even help with a little higher short-term modulus, at least at the high rate of load application/½” diameter deflection per minute, ring strength prescribed by the D2412 testing). They apparently dare not, however, put quite as much fillers like this in pipes that might be used for pressure applications (such as some that apparently allow pipes as thin as SDR 26 for pressure service), as the fillers apparently adversely affect the longer-term properties of the pipe e.g. in stress regression pressure testing.
Now it really gets complicated. Per the ingenious “Iowa formula” for buried pipes developed by Professor Spangler (from my alma mater) decades ago, a distribution of vertical load (in the numerator of the equation) causing deformation of a pipe with walls that move outward against the sidefill soil is resisted by a factored combination of pipe stiffness PLUS soil stiffness (in the denominator of the equation). Aye, “Therein lies the (your) rub”. It would appear the combination of available soil stiffness plus effective (over whatever more inexorable load duration/term the pipes have been buried, not necessarily the higher stiffness values as the result of real rapid loading of a testing machine in a laboratory) pipe stiffness, was not adequate for your local conditions. What you are talking about “on paper” obviously didn’t translate to the real world.
You are not the only one having such problems. In fact, there have been many field problems where very low long-term stiffness plastic pipes have been over-deflected, collapsed, or exhibited other problems related to deformation. Also, multiple specialty contractors have even sprung up around the country who go around and (for an additional cost/price to someone) in effect try to re-round such relatively new pipelines with quite forceful mechanical gizmos (with perhaps also indeterminate long-term effectiveness, or other damage?) While some Engineers have basically tried to make the Contractor responsible for making whatever pipe they specify work, with any manner of soils, depths, and field conditions "work" (and then simply blamed them or others when it didn't after the fact), some authorities have indicated an Engineer's charge with regard to very low, long-term stiffness plastic pipes should go much deeper, so to speak, and in reality not all pipes may be suitable for all conditions. [One is a Past President of ASCE, who authored the piece mentioned e.g. on page 3 of the document at
.]
