Safety Factor (SF)

By Jeff Wright – GBI VP of Technical Services

At Gougeon Brothers, Inc. we are always measuring the ultimate strength of our epoxy products, their adhesive strength, and the strength of the resulting composite laminates (The values measured are often impressive, and the numbers for real-world applications may seem unbelievable). For example, the tensile strength on our Technical Datasheet (TDS) for 105 Epoxy Resin® and 205 Fast Hardener® is 7,900 psi. This means I could easily hang my Chevrolet® Silverado with a full payload using a 1″ x 1″ casting of epoxy! This is true in a perfect application, but our common sense tells us that it is risky. That common sense may be because we understand the need for a safety factor (Sf) in our calculations.

A safety factor is the ratio of the strength of the material to the actual load it will be subjected to in use. An example of a safety factor is comparing the tensile strength of rope to what is often stated as the working load. The rope manufacturer states that when the rope is used in different environments and has aged, it is only safe to use at 20% of the tested breaking strength. In most cases, the rope user values the margin of safety, but it comes at a cost. For example, in sailboat rigging, it is added weight and of course, dollar cost.

Continuing with the rope example, the safety factor for lines used on a racing sailboat in coastal waters with support boats nearby may be much lower than a cruising sailboat used to explore Antarctica. The consequences of failure are part of determining a safety factor. In my experience, the calculations used for a recreational planing powerboat hull bottom may have a safety factor of 4 to accommodate the loss of stiffness from fatigue. The deck may only be 2.5 because a cracked bottom is obviously a bigger problem than a cracked deck.

When using strength data available on WEST SYSTEM® products, keep in mind that the properties listed are the ultimate strength as measured in a testing environment. Although the testing process may not reflect the conditions the epoxy will be exposed to in service, it is important to remember that materials need to be tested in a controlled environment so the results are comparable to tests performed elsewhere. A test lab on the other side of the world can load the sample at the same temperature, load rate, and sample size to compare the results. The data on a TDS is valuable and accurate, just remember it is more than likely the highest strength value achievable.

If your project involves making calculations that may use the strength properties listed on the TDS, or from other test data, the safety factor used should take into consideration some of the potential variables that can affect when an assembly fails:

  • How accurately can the epoxy be metered and mixed? WEST SYSTEM can tolerate common variances when the resin and hardener are measured at the correct ratio, but remember that the TDS lists properties of test samples that were mixed at the exact target ratio.
  • Will a full cure be achieved before the part is put into service? Partially cured epoxy will have much lower properties than those achieved after two weeks at 72°F (22°C).
  • For applications dependent on adhesion, how well can you prepare the surface? Is it difficult to clean or abrade? The adhesion values that we publish are always on properly prepared surfaces.
  • Are you depending on a specific fiber orientation? With fabrics such as a unidirectional fabric, it is important to appreciate the effect of an alignment error. Even a 5° error can have a measurable effect on the strength and stiffness of a unidirectional laminate.
  • Could there be an unanticipated off-axis load in an assembly? In some cases, assuming the load is only placing a load that is 100% in tension on a bonding bracket may cause a failure when the load is slightly off-axis. The difference in the direction of the load could result in a peel load which may cause an unexpected failure.
  • What deflection is there in the assembled parts? Deflection of the assembled parts can generate the peel loads. An example of this would be on a powerboat with very thin hull sides (many of us have tested this by using a closed fist to thump on the hull side to see how much it deflects). When this boat is underway, and the sides deflect outward when the bottom deflects upward, the tabbing on the stiff bulkhead may peel away.
  • What kind of fatigue is expected? If a structure will be subjected to high stresses repeatedly, the effect of fatigue on the strength should be accommodated. This is why older boats often feel “soft”. Excessive deflection has created microcracking resulting in loss of stiffness and strength.

Experienced boat builders have developed strong intuition that enables them to anticipate these types of issues that may result in an unexpected failure. Much of this intuition is based on evaluating the stiffness of a laminate or assembly. In many cases with composite boats, the excessive deflection will provide a warning sign of insufficient strength. If a swim platform is intended to support a Personal Watercraft, it should not have noticeable deflection when a person stands on it, or when standing on a sterndrive the transom should not deflect. If the stiffness is not sufficient, more than likely the structure is not strong enough to withstand the long-term effects of fatigue.

The consequence of failure and variables that are not clearly known should influence the safety factor. For example, if new chainplates are being bonded to a bulkhead, and the loads cannot be accurately determined, the safety factor used to calculate the bonding area should be conservative because the failure consequence is a broken mast. Be sure to use a sufficient safety factor, or work with an expert, to review your plan. A small increase in weight and cost can provide a nice safety factor, which will give you peace of mind when you find yourself in an unplanned situation during use or construction.