Design allowables: why engineers never use the textbook value
Every value on this site is a typical figure — and no real structure is designed to it. How scatter, hot-wet conditions, invisible damage and fastener holes knock a pristine strength down to a fraction before it is used.
The datasheet number is where design starts, not where it ends
Everything in this site's materials database is a typical, room-temperature value with its source. That is exactly right for learning and for a first estimate. But it is worth being blunt about something that separates a student from a practising engineer: no certified structure is ever designed to those numbers.
The value you design to is called an allowable, and it can be a fraction of the pristine textbook figure. Four things drive it down — scatter, environment, damage and holes — and they stack.
Reason 1 — scatter: you design to a statistical floor, not the average
A quoted strength is usually a mean. Make a hundred nominally identical coupons and you get a spread of results, and composites scatter more than metals because failure is governed by the weakest flaw in a heterogeneous microstructure. You cannot design to the average, because half of all parts fall below it.
So aerospace works from statistically-reduced values. A B-basis allowable is the strength that 90% of the population exceeds, with 95% confidence; an A-basis allowable — used where a single failure is catastrophic and there is no load path redundancy — is the value 99% exceed, with 95% confidence. Both sit below the mean, and how far below depends on the scatter, which is why measuring the scatter, not just the average, is half the work.
Reason 2 — environment: hot and wet
Polymer-matrix composites lose properties when they are hot and when they have absorbed moisture, and the certifying condition is usually both at once: elevated-temperature, moisture-saturated — 'hot-wet'. Water plasticises the matrix and lowers its glass transition, so the matrix-dominated properties — transverse strength, shear, and especially compression — fall.
Representative knockdowns for matrix-dominated properties in the hot-wet condition run to roughly 20–40%. Fibre-dominated tension along the fibres is barely affected; the matrix-dominated directions take the hit. This is why a design is checked at the worst corner of its temperature-and-moisture envelope, not at room temperature.
Reason 3 — damage you cannot see
Drop a toolbox on a metal panel and you see a dent. Do it to a carbon laminate and the surface can look perfect while the inside has delaminated — barely visible impact damage, BVID. Because it is invisible, damage-tolerant design assumes it is already there.
The consequence is compression: a laminate with hidden delamination is far weaker in compression than an undamaged one, because the sub-laminates buckle. Compression-after-impact strength is often only about half of the undamaged compression strength, and for many laminates it is the case that actually sizes the part. The design philosophy is usually 'no-growth': the structure must carry limit load with damage present, and that damage must not grow in service.
Reason 4 — holes and joints
Real structures have fastener holes, and a hole in a laminate is far more damaging than in a ductile metal. Metals yield and redistribute stress around a hole; a brittle laminate concentrates it, so an open hole can roughly halve the effective strength. Open-hole tension and, more often, open-hole compression are standard allowables in their own right, and open-hole compression hot-wet is frequently the single governing number for a skin.
Bolted joints add bearing and bypass loads; bonded joints move the problem into the adhesive and the risk of undetected disbonds. Joints are where composite parts are most often sized, and 'just bolt it like metal' is exactly the wrong instinct.
Stacking it up: from 1500 to maybe 500
These factors multiply. Take a unidirectional carbon tape with a pristine, textbook longitudinal strength around 1500–2000 MPa. Put it in a real quasi-isotropic laminate, where only a fraction of the fibres point along the load, then apply a statistical basis, a hot-wet knockdown, and an open-hole-compression-with-damage reduction, and the value an engineer actually designs to can fall to something on the order of a few hundred MPa — often a third or less of the pristine tape number.
Those figures are illustrative, not a formula: the exact chain is program-specific. But the direction and rough magnitude are the point. The gap between 1500 and 500 is the gap between a coupon in a lab and a wing that has to survive twenty years.
How allowables actually get made
Allowables are not looked up — they are earned, through a test campaign built up like a pyramid. Thousands of small coupons at the base establish the statistics and the environmental knockdowns; fewer, larger elements and subcomponents above them capture holes, joints and details; a handful of full-scale articles at the top prove the whole structure. Analysis guides the pyramid, but the certifying evidence is physical test, because the failure modes are too many and too coupled to trust to a model alone.
This is why composite certification is expensive and slow, and why a new material or process means a new campaign. It is also why the provenance on this site matters: a value you can trace is a value you could, in principle, stand behind in that pyramid — a value with no source cannot even enter it.
What this means for using this site
The calculators here give you the mechanics — the ABD matrix, the ply stresses, first-ply and progressive failure — from honest typical properties. That is the right foundation, and it is genuinely what an engineer computes first. Certification then applies allowables on top: the same physics, run with statistically-reduced, environmentally-knocked-down, damage-and-hole-adjusted numbers.
So treat every value on mpolyco as a typical figure for learning and preliminary design — exactly how it is labelled — and treat this article as the reason the real design value is lower, and where that lower number comes from. Understanding that gap is most of what 'master level' means.
1. Why is a certified structure never designed to the mean strength on a datasheet?
2. What is the significance of barely visible impact damage (BVID)?
3. Why does an open hole hurt a composite far more than a ductile metal?