Structural Loads and How They are Determined

For any structure that we build, we decide what forces (loads) the structure will be able to resist during its useful life. We also have to decide what level of service the structure should provide. For example, during a seismic event should the structure show no signs of distress when the load is applied, or should the structure just not collapse? Most of the loads that we design structures for are specified by design codes, but where do the design codes come up with these loads?

There are two basic types of loads. Loads that we have control over and loads that we have no control over. For example, we have control over the weight of the structure itself and the weight of the loads applied to the structure by humans. We do not have control over loads generated by wind, earthquake, or snow. These are controlled by nature. Therefore, the approach for modeling each of these loads is different.

The forces or loads that we design for model loads that will be placed on the structure during the expected life of the structure. We use statistics to determine the level of certainty that a design load will not be exceeded.

For loads that we have control over (say truck weight), we have laws that control what weight and configuration of trucks are permitted on the road. For buildings, we typically post maximum allowable loadings for heavily loaded areas. But, there will still be loads smaller and larger than the design maximum weights. Let?s say we are looking at truck loads. All of the weights of trucks can be plotted like Figure 1.

The plot shows for each truck weight the number of times that that weight will be put on the road. The peak is the truck load that is on the road the most. To determine what weight of truck to design for, we pick a weight that has a small chance of being exceeded, or in terms of the plot above, there is a small area under the curve to the right of the design load (shaded area).

For loads that we do not have control over, we prefer to plot the maximums of these loads for an interval. Say for example, the yearly maximum wind or snow load. If we were to take the maximums and plot them we would have a figure similar to the one below, but it would only plot the yearly maximums. Then based on this plot we can pick a design load that will have a small chance of being exceeded. Figure 2

Again this plot shows how often a yearly maximum occurred during a sampling of years. The peak of the curve shows the yearly maximum snow or wind load that occurred the most.

If only a few years of data are used to plot this figure, then the design load will have more uncertainty than if many years of data are used to plot the figure. This is the approach design codes use to determine the design loads for structures.

The design snow loads are determined such that there is only a two percent probability that they will be exceeded in a given year. Snow loads are typically measured on the ground and then adjusted for the type of structure that they are placed upon.

Wind loads are determined by the fastest mile wind speed (the average speed that a mile of wind passes a point) or the three second gust speed (all wind speeds averaged over three seconds) at a distance of 33 feet above the ground. The design wind speed has a two percent chance of being exceeded in a given year.

Earthquakes are generally looked at over a 50 year period. The design earthquake is usually an earthquake that has a ten percent chance of being exceeded in a 50 year period. Most often we do not have an extensive set of records from past earthquakes for a given area. To determine design loads, geologists map faults in the earth. They then look for past evidence of the movement of these faults. This includes exploration of the earth's surface to show evidence of movement. They develop an estimation of past earthquakes and their probable maximum accelerations. These are then translated into design loads.

It is easy to see that these loads are predictions based on past experience. If we do not have much experience with a given load in a geographical area, then our predicted loads will be less certain than loads in a geographical area where we have more experience.

Many times you hear people talk about a 50 year flood or a 100 year flood. If people talk about a 50 year event, this means the event has a two percent chance of being exceeded in a given year. The 50 years is called the recurrence interval. So if we talk about a 100 year event, this means there is a one percent chance of the event being exceeded in a given year. For most design purposes, the wind and snow design loads are for 50 year events.

Since the forces that we design structures for are unpredictable, any design force can be exceeded during the life of the structure. The question is, what level of risk are we willing to accept so that the structure is economical to build and yet safe? For most loads, we accept the risk that there is a two percent chance the load may be exceeded in a given year. For earthquakes, we use a slightly different approach of accepting a ten percent chance our design earthquake will be exceeded during a 50 year period. We therefore accept that a structure is not wind or earthquake proof. We will never be able to build structures that can resist all of the forces that mother nature inflicts upon us, but with better study of local geology and earthquake behavior, wind, and snow conditions, we will be better able to predict a structure's performance.