The pipe is hydrostatically tested to 1.5 times its design pressure, holds for two hours, passes, and comes out of the trench six months later at the first 90° bend. Nothing was wrong with the pipe, the joints or the backfill. What failed was a force that appears the moment the line is pressurised and that nobody sized, because it does not exist in the pipe wall, does not appear in the stress analysis of the run, and only shows up where the flow changes direction.
Buried GRP is a special case of this problem, and a harder one than ductile iron or steel. A bell-and-spigot gasket joint is a seal, not an anchor: it is designed to keep water in, not to hold a fitting against an axial load measured in hundreds of kilonewtons. GRP's own flexibility, its anisotropic stiffness and its relatively low joint pull-out capacity mean that the thrust restraint is often the single most heavily loaded element in the whole pipeline, and it is almost always the least analysed.
This guide works through the full chain: where the unbalanced force appears, how to calculate it, which pressure case governs, the two ways of resisting it — a concrete block bearing on undisturbed soil, or a restrained length of pipe relying on soil friction — and the calculation, placement and inspection rules that decide whether the design survives the first surge event. All the worked numbers use one buried DN600 PN10 water main so the arithmetic can be followed end to end.
1. Why a Pressurised Pipe Pushes at All
In a straight run of pipe under pressure, the internal pressure acts radially and the axial forces balance out. Each joint is pushed equally from both sides, the net force on the pipe is zero, and the line goes nowhere. Break the symmetry and the balance is destroyed.
A bend is the clearest case. The water arrives along one direction and leaves along another, and the momentum change plus the pressure on the internal walls produces a single resultant force that acts along the line bisecting the angle — straight off the back of the elbow, into the soil. A dead end, a cap or a closed valve is the same problem in its simplest form: pressure acts on the blank face with nothing on the far side to resist it. A tee throws the branch thrust out along the branch. A reducer is a thrust too, because pressure acts on the difference in area between the two sides.
What makes this dangerous in a buried GRP line is that the force is not absorbed gradually by the pipe. If the gasket joints either side of the bend cannot transfer axial load, the entire thrust arrives at the fitting and stays there. The failure mode is a joint pulling apart underground, which appears at the surface as a washed-out trench and a pressure that will not hold — typically on the test, or the first time a pump trips and a check valve slams.
The four locations to enumerate on any buried line are therefore:
- Every horizontal and vertical bend, however small the deflection — a 11.25° joint is still a thrust, just a smaller one.
- Every dead end, cap, blind flange and closed in-line isolation valve. A valve at the end of a phase of work behaves exactly like the cap it is standing in for.
- Every tee, wye, lateral and bifurcation, with the thrust taken out along the branch.
- Every reducer between two diameters, including reducer-to-nozzle transitions on a tank or vessel shell.
2. Calculating the Thrust Force
AWWA M45 (Fiberglass Pipe Design), chapter 7, gives the set of expressions used throughout the GRP industry. The important detail is the area that goes into them: for a gasketed joint where the gasket sits inside the bell, the area is based on the pipe outside diameter at the joint, not the nominal bore or the inside diameter. Sizing on the bore understates the thrust by the ratio of the diameters squared, and it is an error that quietly removes margin from every restraint on the line.
| Fitting | Resultant thrust | Notes |
|---|---|---|
| Bend / elbow | T = 2 · P · A · sin(θ/2) | θ is the deflection angle of the fitting in degrees; a 90° bend gives the largest value for a given area |
| Tee (branch) | T = P · Ab | Ab is the branch pipe joint area; thrust acts along the branch centreline |
| Dead end / cap / closed valve | T = P · A | The full pressure force with nothing balancing it |
| Reducer | T = P · (A1 − A2) | A1 and A2 are the two joint areas |
| Wye / bifurcation | T = 2 · P · A2 · cos(θ/2) − P · A1 | Signed result — the direction depends on the geometry, not just the magnitude |
For the worked example: a DN600 GRP line, PN10, joint outside diameter 630 mm.
- Joint area A = π × 0.630² / 4 = 0.3117 m²
- Design pressure Pdes = 1.00 MPa (10 bar)
- Field integrity test pressure = 1.5 × Pdes = 1.50 MPa (ISO 14692-4)
| Fitting | Thrust at design pressure (kN) | Thrust at 1.5 × test pressure (kN) | Ratio |
|---|---|---|---|
| 90° bend | 440.8 | 661.3 | 1.50 |
| 45° bend | 238.6 | 357.9 | 1.50 |
| 22.5° bend | 121.6 | 182.4 | 1.50 |
| 11.25° bend | 61.1 | 91.7 | 1.50 |
| Dead end / cap / closed valve | 311.7 | 467.6 | 1.50 |
| Reducer DN600 → DN400 | 169.2 | 253.8 | 1.50 |
Two hundred and forty tonnes of unbalanced force from a single 90° bend on a 600 mm water main is not an unusual number, and it is the reason a thrust block is a structural element and not a bag of concrete.
3. Which Pressure Case Governs: Test, Surge or Design
The instinct is to design the restraint at the working pressure. That is wrong in two ways, and both have to be checked.
The test case. ISO 14692-4 requires a field integrity test at 1.5 times the design pressure for process piping, and for pipelines at the lower of 1.5 × Pdes or the pressure that stays inside the occasional-load stress envelope, but never less than 1.25 × Pdes. The test is held for a minimum of two hours with the pressure maintained within ±2.5 %, after being raised gradually over at least 30 minutes. The line is fully restrained and fully pressurised at that moment and the backfill may not be complete, so the restraint should be designed for the test pressure, not the design pressure. On this example, that is a 50 % uplift on every thrust in the table above.
The surge case. A pressure surge arrives as an additional pressure transient, and its magnitude from the Joukowsky relation is Δp = ρ a ΔV, where a is the pressure-wave celerity. GRP is a lower-modulus material than steel, so a is lower — a typical design range is 300–600 m/s against roughly 1000–1200 m/s for a steel main, and the actual value should be taken from the pipe manufacturer rather than assumed. The lower celerity is a genuine advantage, but it is not enough to make surge irrelevant:
| Wave celerity (m/s) | Flow velocity change (m/s) | Surge Δp (kPa) | Peak pressure Pdes + Δp (kPa) | Higher than the 1500 kPa test case? |
|---|---|---|---|---|
| 300 | 0.5 | 150 | 1150 | No — test governs |
| 300 | 1.0 | 300 | 1300 | No — test governs |
| 300 | 2.0 | 600 | 1600 | Yes — surge governs |
| 400 | 1.0 | 400 | 1400 | No — test governs |
| 400 | 2.0 | 800 | 1800 | Yes — surge governs |
| 600 | 1.0 | 600 | 1600 | Yes — surge governs |
| 600 | 2.0 | 1200 | 2200 | Yes — surge governs |
Read that table as a design rule rather than as a table of numbers. At a normal design velocity of about 1 m/s, the hydrostatic test almost always controls, which is convenient because it is the case you must be able to demonstrate on site. But a fast-closing valve, a pump trip against a check valve, or an air pocket collapsing at a high point can produce velocity changes of 2 m/s or more, and at that point the surge overtakes the test case and the restraint is undersized. Either design for the surge explicitly, or control it — slow-closing valves, soft starters, air release valves at high points, and keeping steady-state velocities in the 1.5–3 m/s range rather than pushing for the smallest possible bore.
4. Two Ways to Restrain a GRP Line
There are exactly two mechanisms, and a real line usually uses both.
A thrust block is a mass of concrete placed against the fitting so that the thrust is transferred into undisturbed soil through a bearing face. It does not work by being heavy; it works by having enough bearing area at an allowable soil pressure. AWWA M45 deliberately ignores the dead weight of the block when sizing the bearing area, on the basis that the soil bearing check is the governing one and that including self-weight introduces optimism about the load path.
A restrained or tied length uses the pipe itself as the anchorage. Joints either side of the fitting are positively restrained against axial separation, and the thrust is transferred into the surrounding soil progressively along the pipe through friction between the pipe wall and the embedment. The length required on each leg is set by the friction available per metre.
| Approach | How the load reaches the ground | Requires | Best suited to |
|---|---|---|---|
| Concrete thrust block | Bearing pressure on undisturbed trench wall or floor | Excavation to virgin soil, formwork, concrete cure time before testing, a pipe that can tolerate a rigid hard point | Gasketed joints with no axial restraint capability; competent native soil; accessible fittings |
| Restrained (tied) length | Friction and passive resistance distributed along buried pipe | A joint system capable of transferring axial load; correct embedment and compaction; enough length either side | Confined sites, poor soil, deep trenches, long runs with many small deflections |
For standard GRP bell-and-spigot gasket joints, the restrained-length option is not automatically available — a gasket joint is a seal, and it will separate under axial load unless the specific joint system provides a locking or gripping mechanism. Where a proprietary restrained joint is used, the restraint capacity and the allowable angular deflection per joint are properties of that system and must come from the pipe manufacturer. Where it is not available, the block is the only option, and the design has to work with the soil the site actually has.
5. Thrust Blocks: Bearing Area, Soil Capacity and Placement
The block calculation is a single division. AWWA M45 equation 7-2 gives the required bearing area as:
- Ab = (T × FS) / σ
where T is the resultant thrust, FS is a design factor normally taken as 1.5, and σ is the allowable bearing strength of the soil. Note which soil: the block bears on undisturbed, native ground, and the allowable value is the safe bearing capacity of that material, not of the backfill that was placed around the pipe.
| Soil type | Horizontal bearing strength (kN/m²) | Block area for the 90° bend at test pressure (m²) | Block area for the 45° bend (m²) | Bearing face side, if L = 2H (m) |
|---|---|---|---|---|
| Soft clay | 50 | 19.84 | 10.74 | 6.30 |
| Silt | 75 | 13.23 | 7.16 | 5.14 |
| Sandy silt | 150 | 6.61 | 3.58 | 3.64 |
| Sand | 200 | 4.96 | 2.68 | 3.15 |
| Sandy clay | 300 | 3.31 | 1.79 | 2.57 |
| Hard clay | 430 | 2.31 | 1.25 | 2.15 |
For vertical thrusts acting downward, the safe bearing pressure of the same soils may be taken as twice the horizontal value shown. Upward thrusts — uplift bends — are a different problem entirely and are covered in section 7.
Look at the spread in the first two columns. The same 90° bend needs a 2.3 m² block in hard clay and a 19.8 m² block in soft clay — an eight-fold difference. In soft clay, silt or organic ground, a block of the required size is often impractical, and the design has to switch to a restrained length, to a piled or deep-bearing solution, or to relocating the bend. This is why the geotechnical report is a design input and not a construction formality.
Placement rules decide whether a correctly sized block works. It must bear directly against undisturbed native soil, never against backfill. It must not encase a joint, a bolt, a follower gland or a tie-rod, so that the joint stays accessible for future work. It must be shaped to the fitting so that the load is transferred over the fitting surface rather than concentrated at a corner. And it must fully surround the fitting over its length and circumference so that the fitting cannot rotate or shift off the bearing face.
One useful exemption exists on the piping side rather than the block side. A small branch does not need its connection encased where the branch diameter is 300 mm or less and the header diameter is at least three times the branch diameter — the header wall is stiff enough locally to take the load. Where a branch is eccentric to the header, or is not perpendicular, the branch should be assessed at its longest chord on the header wall, and the exemption usually disappears.
6. Restrained Length: the Soil Friction Arithmetic
When the restraint is transferred to the soil along the pipe, the required length on each leg of a fitting comes from a balance between half the resultant thrust and the friction the pipe can generate per metre. AWWA M45 chapter 7 gives the friction available per unit length as f (2We + Wp + Ww), where:
- We = overburden load per unit length = γsoil × H × D
- Wp = pipe self-weight per unit length
- Ww = weight of contained fluid per unit length
- f = coefficient of friction between the pipe wall and the surrounding soil
- the factor 2 on We accounts for friction mobilised on both the top and the bottom of the pipe
and the length to be tied to each leg of a bend is:
- L = P · A · sin(Δ/2) / [ f · (2We + Wp + Ww) ]
For a bulkhead, a dead end or a tee leg the expression loses the sine term and becomes L = P A / [f (2We + Wp + Ww)], because the full thrust has to be taken out on one leg rather than shared between two.
The inputs for the worked example, with a 1.5 m cover and a granular embedment at 18.8 kN/m³:
| Term | Value | Basis |
|---|---|---|
| We overburden | 17.77 kN/m | 18.8 kN/m³ × 1.5 m cover × 0.63 m OD |
| Wp pipe self-weight | 0.28 kN/m | Typical DN600 SN5000 GRP pipe |
| Ww contained water | 2.77 kN/m | 9.81 kN/m³ × 0.2827 m² internal area |
| 2We + Wp + Ww | 38.59 kN/m | Total vertical load per metre of pipe |
| Coefficient of friction f | Friction available (kN/m) | Restrained length, each leg of a 90° bend (m) | Each leg, 45° bend (m) | Dead end / tee leg (m) | 6 m pipe lengths per leg |
|---|---|---|---|---|---|
| 0.30 | 11.58 | 28.6 | 15.5 | 40.4 | 4.8 |
| 0.40 | 15.43 | 21.4 | 11.6 | 30.3 | 3.6 |
| 0.50 | 19.29 | 17.1 | 9.3 | 24.2 | 2.9 |
Three practical consequences follow.
f is the most sensitive input and the least well defined. AWWA M45 leaves the selection of the pipe-to-soil friction coefficient to the designer, to be justified from the site's own soil conditions and compaction, and it is one of the few numbers in the calculation that a contractor can destroy afterwards by backfilling badly. Taking 0.30 rather than 0.50 changes the required length from 17 m to 29 m per leg on the 90° bend — nearly five pipe lengths instead of three. Where the geotechnical data is thin, design with the lower value and check what it does to the number of joints to be restrained.
A dead end needs roughly 1.4 times the length of a 90° bend. That is a direct consequence of sharing: a bend distributes the resultant over two legs, a dead end has only one. The same logic applies to a wye or a bifurcation, where the thrust acts along one branch.
Count joints, not metres. What the site actually has to do is restrain a whole number of joints on each side of the fitting, so the design output should be "restrain four joints either side", not "restrain 28.6 m". Rounding up costs very little and rounding down is a failure.
The alternative, more conservative in some soils, is the approach in AWWA M23 for PVC, which adds the passive soil resistance against the fitting to the friction and takes Lr = P A tan(Δ/2) × SF / (Fs + ½Rs), where Fs is the pipe-to-soil friction and Rs is the bearing resistance along the pipe. The simpler M45 friction-only form is the conservative choice, because it credits the pipe with nothing beyond the friction it can generate.
7. Vertical and Uplift Bends: When Dead Weight Is All You Have
A vertical bend at the top of a change in grade pushes upward, and friction along a horizontal pipe contributes nothing to resisting an upward force. What resists an uplift bend is the dead weight of the pipe plus its contents plus the soil above it, and the available resistance per metre becomes:
- Resistance per metre = (We + Wp + Ww) × cos(φ − Δ/2)
- L = P · A · sin(Δ/2) / [ (We + Wp + Ww) × cos(φ − Δ/2) ]
Note what changed: the factor of two on the overburden is gone, because there is no friction on a face that is being lifted away from the soil, and the whole load path runs through the weight of the assembly.
| Slope angle φ (°) | Effective resistance (kN/m) | Restrained length, 90° uplift bend (m) |
|---|---|---|
| 0 | 14.72 | 22.5 |
| 15 | 18.03 | 18.3 |
| 30 | 20.11 | 16.4 |
The counter-intuitive result is that the length goes down as the slope steepens. That is real, not an artefact: the term being resisted is the component of thrust along the slope, and on a steeper slope less of the thrust acts in the direction the dead weight resists. What it is not is a licence to reduce the restraint on the basis of slope alone — the calculation assumes the overburden is actually there, properly compacted, and not a trench reinstatement over a void.
Where the dead weight is genuinely insufficient — shallow cover, a lightweight pipe, a large-diameter high-pressure line, or ground that will not carry the load — the options are to increase cover, to add concrete collars or a concrete block to supply the missing mass, to anchor the bend down to a deeper bearing stratum, or to eliminate the vertical deflection by relocating the fitting.
8. The Virtual Anchor and the Axial Stress You Have Just Created
A restrained length is not a free solution, because restraining a flexible pipe against axial movement generates axial stress in the pipe. Buried pipeline design divides the line into two regions. Near a fitting where the thrust has to be taken out, the pipe can move axially — this is the moving pipe. Beyond some distance, friction between the pipe and the soil has fully mobilised and the pipe can no longer move — this is the fixed pipe. The boundary between the two is the virtual anchor, and the distance from the fitting to the virtual anchor is the restrained or tied length. In a typical line, most of the run between fittings is fixed pipe.
A common and incorrect assumption in GRP design is that because bell-and-spigot joints permit axial movement, no axial stress exists anywhere in the line. Field measurements do not support this. Whether a given length is fixed or moving depends on the pipe construction, burial depth, groundwater level, operating temperature and the friction available from the soil — not on the joint type. The safe position is to treat the pipe beyond the virtual anchor as fully restrained and design for the resulting axial stress, both tensile and compressive.
Two mechanisms build that axial stress. The Poisson effect makes a pressurised pipe contract axially when restrained, because a material with a Poisson's ratio around 0.3 that is stretched in the hoop direction wants to shorten in the axial direction; if it cannot shorten, the stress appears as tension. Thermal restraint runs the other way: a pipe installed cold and put into warm service wants to expand, and if it cannot, the restraint produces compression. The two partially offset, which is why a buried line is not simply designed for the sum of both, but they are of different magnitudes and the combination has to be checked against the qualified axial allowable — and against the combined axial-plus-hoop envelope, because GRP is anisotropic and a hoop stress and an axial stress cannot simply be compared with two independent limits.
The axial thermal strain itself is not negligible. GRP is anisotropic, and the axial coefficient of thermal expansion of a filament-wound pipe is generally higher than the hoop value because the reinforcement is predominantly wound in the hoop direction. Typical axial values for a filament-wound polyester pipe fall in the range of roughly 14–30 × 10−6 per °C, so a 100 m restrained run taken from a 15 °C installation temperature to a 35 °C service temperature would want to expand by roughly 40–60 mm if it were free. Buried and restrained, it cannot — and the design has to confirm the resulting compressive stress is acceptable, or provide the expansion provision elsewhere in the system. Above-ground sections, by contrast, must be given expansion joints and loops, and that is a separate design case covered in AWWA M45 chapter 8.
9. Displacement Limits and What a Block Must Not Do
Sizing the block is necessary but not sufficient. A thrust block on a flexible pipe has to be designed so that it does not damage the thing it is protecting, and the industry practice is to bound the displacement rather than only the bearing pressure.
- Limit the fitting displacement to the lesser of 0.5 % of the diameter and 6 mm. For the DN600 example that is 0.5 % × 600 = 3 mm, so 3 mm governs. A block that lets the fitting creep more than that has already taken the joint past the deflection it was designed to accept.
- Restrict radial deformation of the fitting to about 0.5 % of the radius of the adjacent pipe section, so that the block does not ovalise a pipe that is designed to deflect under soil load.
- Do not create a rigid point load. A GRP pipe working as a flexible conduit relies on distributed soil support. A block cast hard against the barrel, rather than shaped to the fitting, produces a local hard point and a stress concentration exactly where the pipe is thinnest.
- Keep the joint accessible. Concrete must not cover a gasket seat, a bolt, a follower gland or a tie-rod. The next person to work on the line will need to reach it.
For a bend at the end of a long restrained length, the block and the tied joints are not alternatives but two parts of one system: the block takes the immediate thrust at the fitting, and the restrained length distributes what remains into the soil over a long section of pipe, so that no single joint sees a load it cannot carry.
10. Joints and Small Deflections on Long-Radius Curves
Many lines change direction not at a fitting but across a run of joints, each deflected by a fraction of a degree. Each of those deflected joints is a small thrust, and they accumulate. The design question is whether the accumulating thrust can be carried by the joints themselves as the pipe curves, or whether the curve needs restraint.
The radius produced by a given angular deflection at each joint follows directly from the geometry of the chord:
| Angular deflection per joint | 6 m pipe length | Resulting curve radius (m) |
|---|---|---|
| 0.5° | 6 m | 688 |
| 1.0° | 6 m | 344 |
| 2.0° | 6 m | 172 |
| 3.0° | 6 m | 115 |
The allowable angular deflection per joint is a property of the specific joint system and must come from the manufacturer; the values above illustrate how quickly the achievable radius tightens. Two constraints apply to any deflected-joint curve. First, the cumulative deflection has to be checked against the seal's capability, because a gasket joint deflected beyond its allowable angle is a leak waiting for the first pressure cycle. Second, the thrust accumulated across a long-radius curve still has to be restrained — on a long horizontal curve the standard treatment is to tie the joints across the curve and let the soil friction along it carry the load, exactly as for a bend.
11. Construction and Inspection: Where the Design Is Won or Lost
A correctly analysed restraint can still fail on site, and it almost always fails in one of five places. Each of them has an inspection point that closes it.
- The block bears on backfill. Inspect the bearing face before the concrete is poured, and photograph it. The face must be cut into undisturbed native soil, with no loose material between the concrete and the ground.
- The concrete has not cured before the test. The hydrostatic test pressurises the line to 1.5 times design when the restraint is often days old. Confirm the cure time and the strength before the test, not afterwards.
- The embedment is not compacted. Every friction and resistance term assumes properly placed and compacted embedment and backfill. Uncompacted backfill can remove most of the friction the design credited.
- The joints were not restrained where the design called for restraint. Walk the restrained lengths before backfilling, count the joints, and record which ones carry restraint hardware. The count in the ground should match the count on the drawing.
- The test pressure at the high point is below the required value. On a line with elevation change the required pressure is achieved at the low point, and the high-point pressure must still be at least 1.25 × Pdes for the integrity test. If it is not, the line has to be split into more than one test section.
Deflection is the other construction check that matters for a flexible pipe and is often skipped because it needs to be measured at the right time: check vertical deflection 24–48 hours after backfilling, at intervals along the trench, and compare against the project limit. For GRP the commonly applied limits are 5 % of diameter for DN 300 and above and 4 % below DN 300, while some specifications are tighter — and the minimum pipe stiffness allowed for a buried application is commonly set at 18 psi (124 kPa) with stiffness classes of SN 1250, 2500, 5000 and 10,000 N/m². A pipe that has deflected beyond its limit at the same time as it is being pushed axially by an unrestrained thrust has two problems interacting, not one.
12. Design Checks Before the Trench Is Closed
The following is the short list worth signing off on every buried GRP line, before the restraint is buried under a reinstatement and becomes unverifiable:
| Check | Input needed | Acceptance |
|---|---|---|
| Thrust enumerated at every fitting | Isometrics with deflection angles, branch sizes, reducers, dead ends and closed valves | Every fitting on the isometric has a thrust value and a restraint type |
| Area used in the thrust calculation | Pipe outside diameter at the joint, not the bore | Thrust computed on joint OD area |
| Governing pressure case | Test pressure per ISO 14692-4 (or AWWA C950 field test) and the surge analysis | Restraint designed for the higher of the two |
| Soil bearing capacity | Geotechnical report, native soil at the block location | σ taken on undisturbed soil, with FS = 1.5 applied |
| Block geometry and placement | Bearing face area, shape, joint access, cure time | Bears on native soil; no joint or bolt encased; displacement limited to the lesser of 0.5 % D and 6 mm |
| Restrained length, if used | Coefficient of friction f, cover depth, soil unit weight, pipe and fluid weight, joint restraint capacity | Design f justified from site data; length rounded up to whole joints |
| Restrained joints counted on site | As-built record against the drawing | Counts match before backfill |
| Axial and combined stress in the fixed pipe | Poisson effect, restrained thermal strain, hoop stress from pressure and soil load | Within the qualified combined axial-plus-hoop envelope |
| Buried deflection | Measured 24–48 h after backfill | Within the project limit (commonly 5 % / 4 %) |
| Hydrostatic test records | Integrity test pressure, hold time, high-point pressure, gauge calibration | Per ISO 14692-4: hold ≥ 2 h within ±2.5 %, high point ≥ 1.25 × Pdes |
13. Sourcing Pipe and Fittings That Match the Restraint Design
The restraint calculation depends on properties that come from the pipe and fitting supplier, and a design is only as good as the data behind it. The four items to obtain in writing before the thrust design is finalised are the pipe outside diameter at the joint, the pressure class and its basis, the joint system's restraint capability and allowable angular deflection, and the wave celerity for surge analysis. For a line carrying potable water, add the approval evidence for the liner system, and for anything beyond water, the resin system's chemical suitability for the actual concentration and temperature.
Where the restraint is by thrust block, the fittings themselves have to be designed for it: the block transfers its load into the fitting, so a fabricated GRP bend or tee with a thin cut-and-mitred wall is not interchangeable with a properly designed fitting of the same nominal size. Ask for the fitting's own pressure rating and its design basis, and ask what the manufacturer's displacement limit for a blocked fitting actually is.
Send us the isometric, the deflection angles, the pipe class, the cover depth and the geotechnical parameters, and we will come back with the thrust at each fitting, the governing pressure case, and a restraint schedule in joints and blocks that a site crew can actually build.
Related reading: how bedding, backfill and ring deflection set the buried design in GRP Pipe Bedding, Backfill and Ring Deflection; the field acceptance regime in FRP Pipe Installation and Pressure Testing; how the 50-year rating is actually established in Long-Term Hydrostatic Strength and HDB; diameter and wall-thickness selection in Selecting FRP Pipe Diameter and Wall Thickness; the overall specification framework in How to Specify GRP/FRP Pipes Correctly; and above-ground support and thermal movement in FRP Pipe Support Spacing and Hanger Design. Product pages: filament-wound FRP water pipe, custom filament-wound GRP process pipe, FRP pressure pipe and the full pipes and fittings range.
Need help selecting FRP products?
Our engineers answer technical questions free of charge.
Ask an Engineer on WhatsApp