Factor of safety is one of the simplest ratios in engineering and one of the most frequently misquoted. It is a strength divided by a stress, and everything interesting about it lies in which strength and which stress. The factor of safety calculator above computes the ratio, reports the margin of safety alongside it, and works backwards to the allowable stress and permissible load implied by whatever factor your specification requires. It is a teaching and preliminary checking tool. It is not a structural design, it carries no engineering stamp, and it does not replace a licensed engineer.
Arb Digital publishes it as part of a free engineering library. It takes its inputs from where the other tools leave off: the cross-sectional area calculator supplies the area, the stress and strain calculator supplies the stress, and this page turns them into a margin. The one thing it deliberately will not do is tell you whether that margin is acceptable, and the reasons for that occupy a section of their own below.
What This Factor of Safety Calculator Does
It divides the strength you enter by the working stress you enter or derive, and reports four things. The factor of safety itself is the raw ratio. The margin of safety is that ratio minus one, expressed as the proportion of headroom above the applied stress — a factor of safety of 2.5 is a margin of safety of 1.5, or 150 percent spare. Aerospace and pressure work tends to quote the margin, structural work tends to quote the factor, and they are two views of the same number.
It then inverts the calculation. Given the factor your specification demands, the allowable stress is the strength divided by that factor, and the permissible load is the allowable stress times the area. The last grid item, utilisation, expresses the applied stress as a percentage of that allowable, which is how most modern design software reports the same information.
How to Use It
- Choose the strength basis. Yield if you care about permanent deformation, ultimate if you care about fracture. The same part gives two very different numbers.
- Enter the strength for your actual material from its specification or mill certificate, at its service temperature. Strength falls with temperature for most metals.
- Give the stress directly, or give a load and an area and let the tool divide.
- Enter the factor your code or specification requires. The tool uses it only to compute the allowable stress and utilisation — it does not check compliance.
- Read the factor and the utilisation together. One tells you the ratio, the other tells you how much of your permitted budget you have spent.
The Formula / How It's Calculated
Factor of safety = strength ÷ applied stress. Margin of safety = factor of safety − 1. Allowable stress = strength ÷ required factor. Permissible load = allowable stress × area. Utilisation = applied stress ÷ allowable stress.
Take the defaults. A member with a 2.0 square inch cross-section carries a 25,000 pound axial load, so the average stress is 12,500 psi. Against a yield strength of 36,000 psi the factor of safety is 36,000 ÷ 12,500 = 2.88, and the margin of safety is 1.88. If the governing specification calls for a factor of 2.0, the allowable stress is 18,000 psi, the permissible load is 36,000 pounds, and the applied stress uses 69 percent of that allowance.
Notice how much information the single figure 2.88 was hiding. It is against yield rather than ultimate, so the part could still be well short of fracture. It is computed on average stress over a plain section, so a bolt hole or a weld toe would raise the real peak. And it says nothing about buckling, fatigue, or whether the 25,000 pound load was the worst case or the typical one. A factor of safety is a summary, and summaries lose things.
What Counts as an Acceptable Factor of Safety
This page will not tell you, and any page that does should be treated with suspicion. The acceptable value is set by the design code that governs your work, by the application, and by the consequences of getting it wrong. A lifting appliance, a pressure vessel, a passenger lift, a building frame, a machine guard and a hobby bracket are all governed differently, and the same physical part can require different factors in different jurisdictions. Codes and standards from bodies such as ASME and the structural specifications published by the American Institute of Steel Construction set the required values in their own domains, and only the code that applies to your work is authoritative for it.
What can be said generally is which way the requirements move. Factors go up when loads are poorly known, when material properties are variable or unverified, when inspection is difficult, when failure is sudden rather than gradual, when people are underneath, and when the consequence of failure extends beyond the part itself. They come down when loads are measured rather than estimated, when materials are certified and tested, when the structure has redundancy, and when failure is detectable before it becomes dangerous. A factor of safety is a statement about uncertainty as much as about strength, and the research programmes at the NIST Materials and Structural Systems Division exist in part because reducing that uncertainty is what allows factors to come down safely.
Yield or Ultimate: The Denominator Changes the Meaning
A factor of safety quoted against yield strength answers a different question from one quoted against ultimate strength. The yield-based factor tells you how far you are from the load that leaves a permanent set in the part. The ultimate-based factor tells you how far you are from the load that breaks it. For ductile structural steels the ultimate strength is substantially higher than the yield strength, so the same part quoted the two ways can differ by fifty percent or more.
Which one matters depends on the failure you care about. A machine frame that must hold alignment fails when it deforms, long before it fractures, so yield governs. A single-use component whose only requirement is not to break might reasonably be assessed against ultimate. Brittle materials complicate this further because they have no meaningful yield point at all, so factors for cast iron, ceramics and concrete in tension are conventionally taken against ultimate and set higher, because brittle failure gives no warning. Always state the basis when you quote a factor; "a factor of safety of 3" without a basis is not a specification.
Where the Factor Went in Modern Codes
Older design practice, still called allowable stress design, does exactly what this page does: it divides material strength by a single factor and compares the result with the calculated stress. Most current structural codes have moved to a limit-state or load-and-resistance-factor approach, which splits that single number in two. Loads are multiplied up by factors that reflect how variable each type of load is, and resistance is multiplied down by a factor that reflects how reliably the member's capacity can be predicted.
The safety has not gone anywhere; it has been distributed to where the uncertainty actually lives. Dead load, which can be calculated accurately from drawings, gets a smaller factor than wind or live load, which cannot. A member whose failure mode is well understood gets a higher resistance factor than one whose is not. The result is a design with a more consistent reliability across load cases than a single global factor could achieve. The practical consequence for anyone using this page is that you cannot compare a factor of safety computed here with a limit-state design and conclude anything — they are different accounting systems for the same risk.
Why the Stress You Divide By Is Usually Optimistic
The average stress on a plain section is the easiest number to compute and the least likely to be the one that fails. Four effects push the real peak above it. Stress concentration at holes, notches, fillets, keyways and weld toes can multiply local stress by two, three or more. Eccentricity adds bending to what you thought was pure tension, and a load offset by even a small distance from the centroid produces a stress gradient across the section — the section modulus calculator handles that side of it. Buckling lets a slender compression member fail at a stress far below yield, which the column buckling calculator addresses, and no factor of safety on yield strength will catch it. Fatigue lets a cyclically loaded part fail after enough cycles at a stress that a static calculation calls perfectly safe.
None of these are exotic. A tension member with a bolt hole in it is the ordinary case, not the special one. Treat a factor of safety computed from average stress as an upper bound on the true margin, and go looking for the local detail that reduces it.
Arb Digital builds free, fast, search-friendly tools that bring technical businesses steady organic traffic. Browse the library or tell us what your customers keep asking you to work out.
Browse Free Tools Talk To Arb DigitalCommon Mistakes to Avoid
- Quoting a factor without its basis — yield and ultimate give very different numbers for the same part, and a bare figure means nothing.
- Dividing by average stress and stopping — holes, fillets, welds and eccentricity all raise the local peak above the average.
- Assuming a static factor covers cyclic loading — fatigue failure happens at stresses a static check passes comfortably.
- Ignoring buckling in compression — a slender member can fail far below yield, and no strength ratio detects it.
- Adopting a factor from another industry — the required value belongs to the code governing your application, not to engineering in general.
Related Free Tools From Arb Digital
Get the section from the cross-sectional area calculator, the stress from the stress and strain calculator, elastic behaviour from the Hooke's law calculator, bending capacity from the section modulus calculator, compression stability from the column buckling calculator and beam reactions from the beam load calculator. Everything else is on the free online tools hub.
Frequently Asked Questions
Divide the material's strength by the stress actually applied. With a 25,000 pound load on a 2.0 square inch section the stress is 12,500 psi, and against a 36,000 psi yield strength the factor of safety is 2.88. Always state whether the strength used was yield or ultimate.
There is no general answer, and this tool deliberately does not offer one. The required value is set by the code governing your application, the consequence of failure, how well the loads and material properties are known, and how easily the part can be inspected. Only the code that applies to your work is authoritative for it.
Margin of safety is the factor of safety minus one. A factor of 2.5 is a margin of 1.5, meaning 150 percent headroom above the applied stress. A margin of zero means the part is exactly at its limit, whereas a factor of zero would mean no strength at all.
Use yield when permanent deformation is the failure you care about, which is the usual case for structures and machine frames. Use ultimate when fracture is the concern, or for brittle materials that have no meaningful yield point. The two give substantially different factors for the same part.
Most current structural codes use a limit-state or load-and-resistance-factor format, which splits the old single factor into load factors applied to each type of load and a resistance factor applied to capacity. The safety is still there, distributed to where the uncertainty is, and the two systems cannot be compared directly.
Not on its own. The ratio is computed from average stress on a plain section, so it misses stress concentration at holes and welds, bending from eccentric load, buckling in slender compression members and fatigue under cyclic loading. Treat it as an upper bound on the true margin.
No, and that is deliberate. Strength depends on the specification, the grade, the product form, the heat treatment and the service temperature, and a value typed from memory into a free web page would be trusted and could be wrong. Take it from the material specification or the mill certificate.
This tool performs a preliminary strength-to-stress comparison for teaching and checking only. It is not a structural or mechanical design, it carries no engineering stamp, and it does not determine whether any factor of safety is acceptable. Required factors, material properties and load cases are set by the code and specification governing your application, and a licensed engineer must design and sign off any load-bearing or safety-critical component.
