top of page

Tips for Trusses


I have designed thousands of truss connections in my career. Below are a few tips I have learned.

  • In general, design both the truss members and the truss connections assuming the joints are pinned*. This will simplify the analysis and the design of both the members and the connections. It is especially important to design the web members of trusses composed of HSS members as axially-loaded members, since the AISC Specification limit states cannot account for end moments at truss members.

  • In order to receive the most favorable pricing and schedule as many of the decisions related to connection configuration as possible should be left to a qualified fabricator, or at least the fabricator should be given the freedom to suggest more economical details.

  • Think outside the box. Box trusses can provide efficiencies in terms of member weight, but in my experience box trusses are almost always replaced with heavier plane trusses when cost is a primary concern.

  • Consider replacing shallow trusses with plate girders, especially when the trusses are heavily loaded. Often to accommodate the loads and the geometry, which can include steep bevels, the truss panels end up being almost entirely filled with gusset plates. Sometimes it is much more economical to provide a solid web plate girder instead.

  • Often trusses are subjected to fewer load cases than connections that are part of the lateral force resisting system. However, for cases where the loading is complex and the forces provided to the connection designer do not satisfy equilibrium, vertical and horizontal transfer forces into the chord should also be provided.

  • Consider the relative depth of the members at a joint. Members with the same nominal depth will have the same dimension between the inside faces of the flanges, but the out-to-out dimension can vary quite a bit. This variation can mean thick fillers, which may have to be developed. It is typical practice to use W14's in large trusses. However, in some instances it may be better to provide a W16 or W18 as a web member to better match the depth and eliminate the need for fillers.

  • Use X-bolts (the threads excluded condition) when appropriate. Often the members and plate thicknesses in long-span trusses and transfer trusses will be such that the thread cannot pass through the shear plane, regardless of field practices. In such case the X-type design values should be used. This can result in a 25% reduction in the number of bolts.

  • If oversized holes are to be used consider providing a couple of standard holes in each joint. This will help to hold the geometry of the truss and ease erection. If the holes end up being misplaced there will only be a couple of holes to be reamed, and the joint will have already been designed as slip critical joints.

  • When using oversized holes where slip is really critical, Turn-of-Nut installation combined with a Class B faying surface will provide the gold standard in terms of reliability. There will be less variation in both the coefficient of friction and the pretension.

  • Take compression loads through direct steel-on-steel bearing as allowed in Section J4.1(b) of the AISC Specification. You will be able to eliminate a lot of bolts. This trick is especially effective for compression chord splices.

  • Members that are to be bolted should be designed based on their net section capacity. This practice is general held to be true, but like most rule-of-thumb it is not without controversy. Steve Hofmeister of Thornton-Tomasetti (formerly of Haven Steel) has stated “…there are many instances where, due to different member actual depths, significant shimming is required between members and connection material. Take for example a heavy truss tension chord splice where the center segment is larger than the next outboard segment. If the outer member size is dictated by an arbitrary …net section capacity, a much larger (and more expensive) section will result.” In some instances there may be little or no savings in up-sizing a member for net sections, but these are the exception and not the rule. In any case the decision should be a conscious one and not just the result of ignorance or apathy.

Large transfer trusses used in a high rise.
Large transfer trusses used in a high rise.



Both the members and the connections for the trusses shown above were designed assuming pinned joints, though obviously there is a good bit of rigidity in these joints. This is common practice in the industry and has been documented in numerous sources, some of which will be provided below.

Gaylord and Gaylord in their text "Design of Steel Structures" (1972) state, "...secondary stresses are usually neglected in design. This is not to say that they are necessarily small but rather that there is sufficient evidence from both analytical studies and experience to indicate that they can safely be neglected..." provided certain precautions are observed. They go on to state that if the ratio of the length of a member to its width in the plane of the truss is 10 or more the secondary stresses will likely be less than 20-30% of the primary (axial) stresses.

Kulicki, Prickett and LeRoy of Modjeski and Masters in Chapter 11 of Brockenbrough and Merritt's "Strucutral Steel Designer's Handook" (2006) state with regard to highway bridges, "Determination of member forces using conventional analysis based on frictionless joint is often adequate when..."

  • The truss is fully triangulated.

  • The workpoints are concentric.

  • Bracing prevents significant out-of-plane distortions.

  • Any camber is induced by lengthening or shortening the members.

  • (Note: This list has been paraphrased.)

In a 2003 Engineering Journal article entitled "Secondary Stresses in Trusses" Shankat Nair, as a representative of the ASCE Committee on Steel Building Structures, states, "the key to proper treatment of secondary stresses in steel trusses is to be consistent between analysis and design. If member forces for design are determined from an analysis that neglects certain stiffness components (such as flexural stiffness of some or all members), stresses corresponding to those stiffness components may be regarded as secondary stresses and may be neglected in design. This approach is consistent with plastic and ultimate design concepts. Limits on secondary stress need to be observed only to guard against local buckling, connection distress, fatigue and other problems which might occur in unusual cases."

With regard to trusses composed of HSS members AISC Design Guide 24 states, "Within the constraints of the limits of applicability in Tables 8-1A and 8-2A, the welded connections within a truss will be semi-rigid (or partially rigid) and branch member stiffnesses will be considerably less than the chord member stiffness. As a result, the actual bending moments in branch members will be very low, and less than would be reflected by a rigidframe analysis." It goes on to state, "If a rigid-frame analysis is carried out, it is recommended that the bending moments generated in the web members be ignored. It is worth noting that the member axial forces produced are generally very similar."

bottom of page