Frequently Asked Questions

Question: I am responsible for a design of what my company calls a “lift beam”, which is a piece of equipment involved in lifting and transporting a maximum load of 34 metric tonnes. The main body of the equipment is based on an ASTM A500 square steel tube. Our procurement team is asking me to allow a supplier to substitute such a manufactured HSS with another that is fabricated from two different flat plates. The plates are each bent into a “U-shape” and then welded together. I think that accepting this substitution is possibly unwise and potentially unsafe, so your comments would be appreciated.

Answer: Square and rectangular HSS are made in North America by three different manufacturing processes: continuous forming with electric resistance welding (ERW) of the longitudinal seam and converting from round to square/rectangular shapes, the most common method, direct forming (or form-square) with ERW of the longitudinal seam, and by submerged arc welding (SAW) of two longitudinal seams joining two C-shaped plates together.

The HSS sizes produced by SAW using two C-shaped plates in the manner that you have described are made to a standard ASTM A1065 (which is similar to ASTM A500), with welding performed to AWS D1.1. These are regularly-accepted HSS and are commonly specified for very big sizes. They are mainly produced by HW Metals (member company of STI), and their website also describes all their available sizes and product grades: https://www.hwmetals.com.

Question: I need to weld an HSS branch member to a thick HSS chord member, but the Specification limits the workable flat width that is available, so how can I make it work? Can I get HSS with a bigger “workable flat” dimension?

Answer: HSS are manufactured in North America by cold-forming and the corner radii are intentionally fairly large to limit the loss of ductility and toughness in the corner regions, promote weldability and to address potential corner cracking. ASTM A500 permits an outside corner radius of up to 3t_nom (where t_nom is the nominal HSS wall thickness) but, on average, the outside corner radius is typically around 2 times the actual wall thickness. In fact, the geometric section properties tabulated for square/rectangular HSS are calculated based on an outside corner radius of 2t_des (where t_des is the design wall thickness, equal to 0.93t_nom). Thus, the corner radii can be relatively large for thick-walled HSS members, but it is inadvisable to try to source cold-formed HSS with very tight corners because of the problems with mechanical properties that this might entail.

For determining the cross-sectional compactness of a square/rectangular HSS, one needs an estimate of the largest flat dimension that may possibly occur (to determine b/t and h/t), and for this the outside corner radii are taken as 1.5t_des. For detailing, one needs an estimate of the smallest flat dimension that may possibly occur, and for this the tabulated dimensions (in Tables 1-11 and 1-12 of the AISC Manual) are calculated using outside corner radii of 2.25t_nom. This means that for making a connection to an HSS member one can rely on a flat width = (B – 4.5t_nom), where B is the outside width of the HSS. This does limit the space for fillet welding all around a branch member, if the fillet weld leg along the branch longitudinal side is to land on the flat of the HSS “thru member”. However, welded joints can still be made – even up to matched-width HSS connections – by using partial joint penetration (PJP) flare-bevel welds along the branch longitudinal sides.

Question: Table J2.5 of the AISC 360-10 Specification, for fillet welds loaded in shear, indicates that two limit states are pertinent – failure of the weld metal and failure of the base metal. Does that mean that weld design in HSS joints is based on checking weld failure through the effective throat, plus base metal shearing along the fusion face?

Answer: No. Fillet weld strength is determined on the basis of failure of the weld metal, in shear, along the weld effective throat. The note in Table J2.5 about shear failure on the base metal being governed by Section J4 is meant as a reminder that shear through the thickness of the connected wall may need to be checked. Hence, for a stepped HSS-to-HSS T-connection, fillet-welded all around, the weld is only proportioned on the basis of failure along the effective throat. This requirement is the same in the current Canadian standard for steel structures too (CSA S16-09). Note that the prior Canadian standard (S16-01) did require a check for shear failure along the fusion face. All of the foregoing is based on weld filler metal with a strength level equal to, or less than, “matching filler metal”. If over-matching filler metal is used then it would still be prudent to perform an additional check for shear failure along the fusion face.

Question: Are different AWS D1.1 prequalified welds required around the perimeter of a 45° miter butt joint used to connect two pieces of HSS at 90 degrees? Or, can the same weld be used around the entire perimeter? In either case, what is the prequalified weld typically used? A partial-joint-penetration groove weld?

Answer: In direct answer to your question, edge preparation (beveling) of the HSS would be required along 3 of the 4 edges of the miter joint, in order to successfully accomplish PJP welds along those 3 edges.

The capacity of such welds is limited, and the welds themselves (to matched box sections) need to be made very carefully. For such miter joints, especially if there is reasonable applied loading on the HSS members, it is recommended that the two HSS each be separately welded (usually by fillet welds) to a 45° stiffening plate. This type of “knee connection” is shown and discussed on pages 67 to 69 of CIDECT Design Guide No. 3 – “Design Guide for Rectangular Hollow Section (RHS) Joints under Predominantly Static Loading”, 2nd. edition, 2009.

Question: How do you account for residual stresses in welded connections to HSS members?

Answer: The residual stresses due to welding of HSS connections are inherent in all welded HSS construction and the effects are already included in connection design equations.

Question: I am designing a structure here in Canada and I am considering the use of steel hollow sections. The structure is exposed to extreme temperature and reversal of forces. As such, one requirement that was imposed by the client was that all structural steel shall be Grade 350WT comply with Charpy Test requirement of 20 Joules at -20 deg. Celsius. Grade 350WT has a minimum yield strength of 350 MPa (or 50 ksi) and tensile strength of 450 to 620 MPa (or 65 to 90 ksi). Appreciate if you could advise whether such HSS material is readily available in North America.

Answer: The specification of a notch tough steel for a cold, dynamic application is very wise, and CSA G40.20-13/G40.21-13 Grade 350 WT Category 2 seems to have been specified. (Category 2 corresponds to 27J at -20°C, whereas you have actually stated 20J, which is a slightly lesser requirement. Normally 27J is the stipulation in Canada, per Table 9(b) of the CSA “Structural Quality steel” standard). Grade 350W/50W is readily available in Canada (although ASTM A500 comprises around half of the national market), but the notch-tough WT grade may need to be ordered from a tube producer as it is a special requirement. 

There is another grade of HSS now produced in North America since mid-2013: ASTM A1085. This has superior structural properties to ASTM A500 and could be substituted for CSA G40.20/G40.21 Grade 350 WT Category 1. (Category 1 has a minimum Charpy V-Notch toughness of 27J at 0°C, whereas ASTM A1085 has a minimum Charpy V-Notch toughness of 34J at 4°C). If a HSS supplier suggests using ASTM A1085 material then you would have to invoke a supplement S2 add-on, which allows a special, more stringent, Charpy V-Notch toughness to be specified.  However, since your project is in Canada you may as well stick with CSA G40.20/G40.21 Grade 350 WT Category 2 material.

Question: Where can I best obtain the axial compressive capacities of concrete-filled HSS columns?

Answer: The current design of composite members is prescribed by Chapter I of ANSI/AISC 360, which is a free download from www.aisc.org. Further information for designers is available from the current AISC “Steel Construction Manual”, also available from the AISC website. In STI Design Guide Volume 2, tables  outlining available strengths in axial compression (for both LRFD and ASD) for square, rectangular and round HSS to ASTM A500 Grade C, and ASTM A1085, filled with 4-ksi normal density concrete.

Question: I have designed a tube header for a door that is taking both horizontal wind and also a vertical load at the midspan of the header. My customer is now requesting to allow for 2″ diameter holes @ 2′-0″ o.c. along the top and side of the tubing. I am told this is for insulation. Taking out this much section property concerns me and I am not sure quite how to check this. What checks would you recommend? Do you think I should be concerned? I am thinking first off that they should space the holes apart further. Suggestions?

Answer: The implications of two-inch diameter holes, at two-foot centers, will very much depend on the size of your HSS (presumably square or rectangular) and the magnitude of the applied loadings.

If holes are being put in both the side and the top of the HSS it would seem prudent to offset them, if possible, so that two holes do not occur at the same cross-section. It seems like you have a bi-axial bending situation under wind + gravity loads. You could calculate the reduced cross-sectional properties of the HSS (with portions of steel removed) and use these properties at the critical cross-sections (points of maximum bending moment and maximum shear force). You should be checking the reduced bending capacity, the shear capacity (which is not normally critical but it may be, if parts of the web are removed) and the deflection.

Question: I can’t find any information about unbraced lengths for HSS beams. When using HSS sections as beams, what moment can a beam resist given a certain bracing spacing, or vice versa, what spacing the bracing needs to be in order to resist a certain moment?

Answer: The user note in AISC 360 specification section F7 states that square and rectangular sections are not usually subject to lateral torsional buckling. The commentary goes into more detail about why. So, you can use the equations in F7 for designing HSS sections as beams.

Question: What is the assumed corner radius for rectangular HSS sections? The value I calculated based on the tabulated b/t ratios and workable flat dimensions do not match up.

Answer: You are correct in that the corner radius assumed varies based on the use of the radius. HSS members are produced to an ASTM standard and the standards do have tolerances for the corner radius. There is a maximum tolerance of 3 times the nominal wall thickness (3t_nom) for ASTM A500 and no minimum. For ASTM A1085 where the wall thickness is less than 0.4 in., the minimum radius is 1.6t. The minimum is 1.8t for thicker walls. In both cases the maximum is 3t. The industry standard corner radius is about 2 times the actual thickness of the HSS walls. Section properties used for the calculation of capacities are based on the industry standard. The section property tables (Table 1-11 and 1-12) in the AISC Steel Construction Manual are based on a corner radius of 2 times the design thickness (which is not necessarily equal to the nominal thickness), as discussed on page 1-5 of the 14th Edition Manual. The STI Design Manual: Volume 1 also contains HSS section properties and uses the same corner radius as indicated in the ‘HSS Properties’ section of the manual. When calculating the width-to-thickness ratios used to classify HSS sections for local buckling, it is conservative to use a smaller value for the corner radius, resulting in larger element widths. AISC Specification Section B1.4b(d) defines the width, b, as the outside dimension minus three times the design wall thickness. Therefore, the corner radius is taken as 1.5 times the design wall thickness for these calculations. The values tabulated for ‘b/t’ and ‘h/t’ in the AISC Steel Construction Manual and the STI Design Manual are based on 1.5 times the design wall thickness. Both the AISC Steel Construction Manual and the STI Design Manual indicate that the workable flat dimensions are based on a corner radius of 2.25 times the nominal thickness of the member (since it is a detailing dimension) accurate to the nearest 1/16 inches. However, for large ASTM A1065 sections, the STI Design Manual uses 3 times the design thickness, which correlates with the industry standard for production of A1065.

Question: I am a structural engineer out of Chicago, and I am looking for information on HSS tube products. More specifically, I am trying to find some definitive section properties for small tubes made from gauge-thickness steel. The projects that I work on typically employ HSS 6×6 and smaller sections (down to 1×1 and including rectangular shapes). We have used sections with wall thicknesses as small as 14 gauge. Thus far, we have been making educated guesses for dimensions and section properties, but would really like to have some concrete information.

Answer: The Steel Tube Institute has published a brochure that includes small sizes outside of those published in the AISC Steel Construction Manual.

Question: I need the rotated section properties for square HSS. The working cross-section is thus diamond shaped. Do you publish these section properties?

Answer: I am not aware of any published section property data for HSS (or open profile shapes for that matter) on sections with transformed axes. 

The moment of inertia (I) property is given by the AISC 15th Edition Steel Construction Manual Table 17-27 as I = I_xsin²ϴ + I_ycos²ϴ, which for a 45⁰ angle would result in I_rotated = I_x = I_y. For other section properties, as necessary, one might even consider calculating based on a sharp-cornered section; i.e. do not consider the effect of the corner radii. Of course, for the 45°-rotated section the cross-sectional area (A) and radius of gyration (r) also remain the same as for the un-rotated section.

Question: We are looking at using an HSS material with a 70 ksi minimum yield strength that is not currently and accepted material in AISC 360. What do we need to be aware of pertaining to the design and are there other codes that are applicable to this material?

Answer: Some European manufacturers currently are able to provide square/rectangular HSS of very high yield strength. In laboratory tests that have been performed, to date, on high-strength HSS there have been no adverse properties detected, except you need to take account of the high yield-to-ultimate strength ratio (F_y /F_u) in connection design (because reasonable ductility levels are presumed, and the elongation of this product is lower than for ASTM A500 product) and the higher degree of connection deformations.

The design rules in Chapter K of AISC 360 are based on the 2nd. Edition, 1989, IIW recommendations [1], which were valid for HSS up to a nominal yield strength of 360 MPa [52 ksi]. The more recent, 3rd. edition, 2012, IIW recommendations [2] – which are very similar to the 2nd. Edition design rules – have extended the range of application to HSS with a nominal yield strength up to 460 MPa [67 ksi]. However, whenever the HSS nominal yield strength exceeds 355 MPa [51.5 ksi], as in your case, connection design must be performed as follows [2]:

(i) The design yield stress should be taken as the lower of the nominal F_y value and 0.8 of the nominal F_u value.

(ii) All connection resistances, determined using the published formulas, need to be reduced by 10%.

Thus, this philosophy could be applied to design of your high-strength HSS connections, with a yield strength cap of 67 ksi. So, although the AISC 360 Specification does not specifically reference these high-strength HSS this material should be acceptable, with the provisions noted above.

The minimum guaranteed yield strength (F_y ) can be fully utilized in the design of compression members.

References

[1] IIW, 1989. “Design Recommendations for Hollow Section Joints – Predominantly Statically Loaded”, 2nd. Edition, IIW Doc. XV-701-89, International Institute of Welding, Paris, France.

[2] IIW, 2012. “Static Design Procedure for Welded Hollow Section Joints – Recommendations”, 3rd. edition, IIW Doc. XV-1402-12, International Institute of Welding, Paris, France.

Question: For one of my projects, a designer has specified some HSS frames of A500 Grade 50ksi yield strength. The supplier has proposed an A500 Grade B but this type of steel, as per the ASTM A500 standard, has only a 46ksi yield strength. So I asked him to find A500 Grade C and now he’s responding that when steel suppliers purchase HSS tubes from the mill, both material types are retrieved from the same stockpile, essentially saying that they are the same.

Is that true? Can we go with A500 GrB instead of A500 GrC?

Answer: No. If a designer has specified ASTM A500 Grade C HSS then you must use that grade. It is true that all STI manufacturers in North America will dual-certify their HSS, meaning that it meets both ASTM A500 Grades B and C. If the manufacturer’s test certificate states this, then this product is acceptable for use with both Grade B and Grade C designs. If the manufacturer’s test certificate only states that the product meets ASTM A500 Grade B then it can only be used for Grade B designs.

Question: Do you have any advice for strengthening an existing HSS square column by pumping grout to fill it in, or by welding plates on each side? I do not want to use FRP strengthening methods. The tube is an 18’ tall HSS 6x6x1/4 column.

Answer: I think you have identified the three main methods available for strengthening. Given that this is an existing column, the insertion of grout inside the HSS would be difficult. An input hole would have to be made at the bottom and an outlet hole at the top, and the bottom one would have to include a tap for the pumping apparatus too. I am not aware of such concrete- or grout-filling done on an HSS column after it is in service. Welding on steel plates to the HSS column is a straight-forward procedure, especially as the column is not circular, and the increased column section properties can be calculated with this addition of steel. Depending on the increase in column strength sought, you might want to consider welding plates on just two sides, as this involves much less site welding. Plates on two sides would effectively create a member with different properties about the two axes (i.e. a weak and a strong axis), but the column buckling strength would still be raised because the moment of inertia (I), about both axes, increases. The strengthening of HSS compression members by the addition of welded steel plates has certainly been done before.

Question: I am installing numerous steel carports at California public schools which employ HSS tubing. The California school building authority (DSA) gives us greater design flexibility if we can prove that our carports are built of noncombustible materials per ASTM E136 (furnace test). Can you help me prove that HSS sections are noncombustible per California Building Code 703.4, which references the materials passing ASTM E136?

Answer: I don’t have any information (such as a test report) in this regard, but the ASTM E136 standard deals with assessing the combustibility of building materials, whereby a small sample is put into a small furnace.

The E136 test would give results similar to what ULC has in their Canadian S114 standard, entitled “Standard Method of Test for Determination of Non-Combustibility in Building Materials”. In this regard the steel industry has a ULC letter (available upon request) which declares that steel is non-combustible, and this might be of use. This letter lists some of the criteria in the test, one of which is a 750°C temperature limit. This is the same temperature which ASTM E136-11: Standard Test Method for Behavior of Materials in a Vertical Tube Furnace at 750°C uses. ULC and ASTM do have some differences; e.g., the time of exposure at 750°C – 15 minutes in ULC S114 and 30 minutes (clause 4.1) of ASTM E136.

Question: We are trying to find design criteria for using HSS webbing with flattened ends in long-span trusses. We need to flatten the ends to fit between two angles, at the top and bottom chords of the truss. I know that this is done, but I can’t seem to find design criteria for it. I used to use criteria dating from 1991 by CIDECT in their “Design Guide for Circular Hollow Section (CHS) Joints under Predominantly Static Loadings.” Our trusses range from 90’ to 125’ long. We are estimating about 1000 trusses to be built per year. So this will be a lot of material working, and we would like to be as efficient as possible. Do you know of any additional resources related to flattened ends?

Answer: Flattening the ends of round HSS (or pipes) is used in lightly-loaded trusses, and can be used in conjunction with pairs of separated angles for each chord, as you have described. The flattened web member end is inserted between two angle legs and welded to them. Most guidance that is available pertains to the flattening process itself, or the connections when flattened ends are welded to round or square HSS chord members.

The CIDECT Design Guide (No.1) which you have accessed has some discussion on this topic. The most recent (2nd. Edition, 2008) version of this CIDECT Design Guide covers the topic on pages 68 to 70. This information is in accord with the book, “Hollow Structural Section Connections and Trusses” by J.A. Packer and J.E. Henderson, 2nd. Edition, 1997, published by the Canadian Institute of Steel Construction, Toronto, ON (www.cisc-icca.ca).

This latter publication (not available in electronic format) does give broader treatment (pages 148 to 155), but still related to HSS as the chord members. CIDECT Design Guide No.7 (dealing with “Fabrication, Assembly and Erection of Hollow Section Structures”) has some information on the flattening process itself.

Question: I’m working on a project where we have both HSS beams and HSS columns for an exterior frame. We’re planning on providing weep/drain holes in the bottoms of the HSS beams and at the bottom of the HSS columns to allow for drainage from condensation/rain water since this is an exterior frame completely exposed to the elements, but I’ve been unable to find any literature documenting what size and spacing such weep holes would typically be provided. Any guidance that you could provide on this matter would be greatly appreciated.

Answer: Recommended practice (given by Stelco brochures dating to the 1980s) for water drain holes in HSS columns is shown below, and I imagine similar sized holes would be appropriate in beams. Try to keep the column drain hole as close to the HSS bottom as possible/practical.

Question: I am using ASTM A500 tubing in a design and I would like to know if you have any standards for bending it. In particular, what are the minimum bend radii and where should I dimension it to – the centerline or the inside radius? If you do not have any standards, can you direct me to a place that does?

Answer: HSS can be bent or rolled into curved lengths using either cold or hot processes. Cold working is relatively inexpensive and readily available, but smaller bend radii are obtainable with hot bending techniques. Kennedy [1] performed experimental research on cold bending of HSS and found that, at the (tight) bending limit, two distinct forms of HSS deformation were possible: (i) inward bowing of the compression face of square/rectangular HSS, and (ii) outward bulging of a side face of square/rectangular HSS. Relatively thin-walled HSS exhibit early compression face bowing, while thicker-walled HSS tend towards earlier side wall bulging. Minimum cold-bending radii that limit the amount of wall distortion (to 1% or 2%) are given in Section 8.1.3.1 of Packer and Henderson [2], for both square and rectangular HSS with the latter bent about both axes. Such recommendations are still tentative, however, as bending results do show some dependence upon the rolling machines used, the actions of the operator, and the level of HSS wall support during bending. Since the HSS has been cold-worked during (cold) bending, caution should be exercised upon galvanizing unless the material has been stress-relieved.

[1] Kennedy, J.B. 1988. “Minimum Bending Radii for Square & Rectangular Hollow Sections (3-Roller Cold-Bending)”, CIDECT Report 11C-88/14-E, University of Windsor, Windsor, Canada.

[2] Packer, J.A. and Henderson, J.E. 1977. “Hollow Structural Section Connection and Trusses – A Design Guide”, 2nd Edition, Canadian Institute of Steel Construction, Toronto, Canada.

Question: I would like to inquire about the feasibility of using a 12”x12” structural tube to encase an existing concrete column. Is it possible to use two halves of a tubular section and weld in the field? If so, what are the pros and cons?

Answer: It is possible to encase a reinforced concrete column with a square hollow structural section (HSS). Presumably your existing concrete column is also square, and somewhat smaller in dimensions. You could not expect the HSS casing to fit snugly onto the existing concrete column because the inside hole of an HSS is not square, due to the rounded corners, which have an inside radius of approximately t (the design thickness of the HSS). Thus, the HSS will need to be “oversized.” The HSS will then have to be cut longitudinally into two C-shaped sections and backing shop-welded along the entire length to the inside edges of two tips of one of the C-shapes. Then the two Cs can be put around the concrete column and tacked together at the exposed backing. Then, field groove welding along the full length of the column, on two sides would take place. Finally, the gap between the HSS and the concrete column should be filled with a non-shrink concrete grout to achieve integral action between the HSS and the original concrete column. To install this grout, pumping from the bottom via a pre-installed access hole in the steel casing would be optimal (as is done for concrete-filled HSS columns). If you are looking for eventual behavior as a composite column, bear in mind that there is little confinement pressure exerted on concrete infill within a square (or rectangular) HSS composite column. For a circular HSS composite column, on the other hand, it is possible to achieve considerable concrete confinement action – and thus there has been rehabilitation of circular reinforced concrete members using this steel encasement technique described above.

Question: Should we as SER’s be specifying vent hole size/locations for galvanizing, or should this be in the fabricator’s scope? Also, we frequently specify end/cap plates that are seal welded. The presumption is that we are making the tube tight to moisture/environmental conditions, but realistically we’ll get condensation. What is recommended practice for putting permanent vent holes? (i.e., (1) 1” diameter hole per 30’???)

Answer: In my opinion, yes, we as the EOR should be specifying vent hole locations. Fabricators may not be aware of structural and/or architectural considerations that may be sensitive to the location of such a hole. If you don’t really care where the hole goes, then I suppose you could leave it up to the fabricator, but personally I tend to err on the safe side. I haven’t seen a vent hole being specified in relation to its length. Usually specifying one at each end of a member should be sufficient. If the member will truly be exposed to moisture you might consider plugging vent holes to prevent water from entering the cavity of the tube; although even in that case, I do think it is prudent to leave a small hole at the base of a column to allow any condensation or water in the member to escape.

Question: Our company is designing a utility line crossing, to span a floodplain and stream. We will be suspending 300′ of 10″ HDP water line with a uniform load of 47 lb/ft (water & pipe). We would like to keep support piles to a minimum because of wetland disturbances. Our options are either round steel casing or square tubing, somewhere in the range of 14″ to 16″ in diameter. We would then run the waterline through the steel to the opposite side of the stream. The question is: what is the maximum span possible given a specific size and thickness? 50′ would be an ideal span given the stream is 40′ wide.

Answer: The design of a load-bearing HSS beam to span 50 feet should be performed by a qualified and licensed structural engineer. However, “Available Flexural Strength” tables (for use with LRFD or ASD) for HSS beams, designed in accordance with AISC 360-16 Sections F7 and F8, are given in the AISC “Steel Construction Manual”, 15th. Edition. These tables cover rectangular HSS bent about its strong axis (Table 3-12), Table 3-13 (square HSS), and Table 3-14 (round HSS), all using ASTM A500 Grade C material. For such a long-span flexural member, note that deflection should also be considered, as well as strength.

Question: Would a HSS5.5×5.5 telescope into a HSS6″x6″x1/4″, for on-site construction assembly?

Answer: Square and rectangular HSS are not well-suited for telescoping applications. Although the HSS6x6x1/4 member has a theoretical distance (because the wall thickness may actually be greater than nominal, although rare, due to permitted tolerances) of 5.5” between the inside walls, the interior clearance is confounded by two things:

(I) The inside corners of the HSS have a typical radius of 1.0t and the outside corners 2.0t (where t is the actual wall thickness), but these corner radii are highly variable – both above and below these values.  Furthermore the four inside corners of the same HSS can all have different inside corner radii.  The inside / outside corner radii may also vary between HSS manufacturers.

(II) There is a longitudinal weld bead protruding towards the inside of the HSS, along the whole length, which will foul any inserted member. Most manufacturers can remove this weld flash, but it must be ordered that way at the time of manufacture (and costs extra to do so) which means that the HSS quantity must justify doing this.  This is sometimes done for round HSS, which is the type commonly used for telescoping applications.

Question: I have an HSS 203 x 76 x 3.09 mm (Class 4 cross-section) with a known F_y of 426 MPa (from experimental coupon tests). To obtain the axial compressive capacity, I referred to CSA S16-09 Clause 13.3.5 dealing with “Members in compression subjected to elastic local buckling.” I evaluated the capacity based on the two methods given – (a) and (b) – but I get two vastly different results. Which one is more correct?

Answer: Initially, it should be reiterated that a measured yield stress is not permitted to be used in structural designs – the minimum guaranteed (or nominal) yield stress must be used.

As implied in this clause of the Canadian code, both methods for handling slender cross-sections are permissible:

Method (a) is based on the determination of an effective area, calculated using reduced element widths. This can be used for square and rectangular HSS, and box sections, but not for round HSS or pipes.

Method (b) is based on computing an effective yield stress for the whole cross-section and can be used for any shape of cross-section.

The two methods do not give consistent results for square/rectangular HSS. If the member is stocky with respect to overall flexural buckling (i.e. has a low KL/r) and fails by yielding, computing an effective (reduced) area – method (a) – whose elements satisfy the local buckling requirement, gives the greater buckling capacity. If the member is slender with respect to overall flexural buckling (i.e. has a high KL/r) and fails by Euler buckling, computing an effective (reduced) yield stress – method (b) – that satisfies the local buckling requirement, gives the greater buckling capacity.

In summary, it is safe to compute the member compressive strength of a rectangular HSS by using both methods and to take the larger of the two.

Question: I am an architect interested in using rectangular steel tube for a project – but we would like to minimize the corner radius for aesthetic reasons. What are my options?

Answer: The corner radius of Hollow Structural Sections (HSS) is not something that can be manipulated to order, and is part of the existing manufacturing paradigm of numerous HSS producers in North America. The typical outside corner radius of North American HSS is 2t, where t is the wall thickness of the as-produced HSS, to the standard ASTM A500. Based on such typical outside corner radii, the width of the “flat” is relatively predictable for steel fabricators and it is also built into the wall slenderness design checks (for section compactness) in AISC 360. Even the existing corner radii are relatively “tight”, considering that HSS are manufactured by cold-forming. Even smaller corner radii would induce excessive amounts of cold-working in the corners. If you really require sharp corners for an aesthetic reason, it is always possible to prescribe box sections, which are made from four plates welded together. However, this is a much more expensive solution and the appearance of the corner welds may still be a concern to you. To minimize the size of the outside corner radius of manufactured HSS the wall thickness of the HSS needs to be low (due to the 2t relationship mentioned above).

Question: The flexural design provisions for HSS in AISC 360-10 Chapter F are based on F_yZ; i.e. the plastic modulus is used. However, the calculation of U in Chapter K is based on F_yS; i.e. the elastic section modulus is used. [The actual U-equation is P_r/(A_gF_c) + M_r/(SF_c)]. (i) Why is there this difference? (ii) This could lead to U>1 for highly optimized members. Should one use a U>1.0, or if U is calculated > 1.0, then should U=1.0 be used? (iii) Is additional modification necessary for biaxial bending?

Answer: (i) The expression for U in Chapter K derives from international practice, and is mated to the rest of the adopted formulas. (Nevertheless, it is always conservative to use the elastic section modulus, rather than the plastic section modulus, for all flexural limit states and compact/non-compact sections). The AISC Specification committee could look at changing the denominator in the moment utilization term to M_n to bring this into line with Section F7. (ii) U is intended to be the chord utilization, based on axial stress + bending stress, on the connecting surface. Thus, I think it should philosophically be limited to 1.0. (iii) Once out-of-plane bending occurs, the stresses are no longer all additive like axial + bending normal stresses, so this is certainly in the realm of “engineering judgment”.

Question: A colleague, upon checking calculations for shear in a tube, asked me if I know of a reference for the 0.93 factor he sees being applied. Do you know of this?

Answer: Cold-formed HSS in the US is manufactured to the standard ASTM A500. This production standard has no tolerance on mass/weight and allows a wall thickness under-tolerance of -10%. Manufacturers have thus become accustomed to producing the finished product with a wall thickness close to 0.90 times the nominal thickness (as this saves on the amount of steel used). However, the section properties (e.g., A, I, r, S, Z) used in structural calculations – for safety – need to be determined on the basis of the actual wall thickness, not the nominal. After consideration of all structural reliability issues, the American Institute of Steel Construction (AISC) determined that a wall “design thickness” of 0.93 times the nominal thickness would be appropriate, and would place HSS on a similar structural reliability level as other hot-formed open sections and then enable the use of common resistance/safety factors. Naturally, the design wall thickness of 0.93*t_nominal applies to all structural calculations, including those for available shear.

Question: The limits of validity in Chapter K of AISC 360-10 indicate a branch angle of greater than or equal to 30 degrees. Does this apply if the truss is curved?

Answer: The limits of validity would apply to all trusses, regardless of whether the chords were straight or curved. This restriction was introduced as a practical measure to produce proper welds, especially in the K-connection heel position. Thus, it is based on welding access, and not necessarily a limit on the validity of the design equations. This angle restriction was removed in AISC 360-16 Chapter K, and replaced by a User Note pointing the designer to the fact that low branch member angles must be confirmed with the fabricator to ensure that acceptable welding can be performed to meet all standards.

Question: Is there a specification that clearly defines where the seam must be located on square and rectangular steel tubing? We recently received some tubing with the weld seam directly on the corner.

Answer: Section 6.3 of ASTM A500-13 intends that the weld seam not lie anywhere within the curved region; i.e. not past the tangent point to the flat.

Question: The commentary to AISC 360-10 Section K1.3 says that, “If a longitudinal plate-to-rectangular HSS connection is made by passing the plate through a slot in the HSS and then welding the plate to both the front and back HSS faces to form a “through-plate connection”, the nominal strength can be taken as twice that given by Equation K1-12, and is given by Equation K1-13.” Can the same thing be said of equation K1-2? Can you double the nominal strength if the plate is a “through-plate connection” on a round HSS?

Answer: Yes – providing the through-plate is longitudinal to the round HSS member, and hence the limit state is plastification of the HSS, a through-plate connection should give you double the capacity of that given by AISC 360-10 equation (K1-2), because that equation is conservative (shown in recent research by Voth and Packer). An earlier paper on this topic (Willibald et al., ISTS11 conference, 2006) showed comparative laboratory tests where a single longitudinal through-plate connection had a capacity equal to 1.6 times a longitudinal branch-plate connection. However, equation (K1-2) under-predicts the capacity of the longitudinal branch plate connection significantly.

Question: I am designing a moment frame using HSS tubes for a seismic region (California). The plan checker is refusing to accept it as an “Intermediate Moment Frame” (IMF) because it does not comply with AISC 341. Do you have any information regarding the use of HSS moment frames and how they can be classified for seismic design as “ordinary”, “intermediate” or “special” (OMF, IMF SMF)?

Answer: A seismic force-resisting system, using moment-resisting frames with HSS columns, and either wide flange beams or HSS beams, can be designed to be an Ordinary Moment Frame (OMF), as defined by AISC 341 Chapter E, with some attention to the connection details and reinforcement. However, in order to be classified as an Intermediate Moment Frame (IMF) or a Special Moment Frame (SMF), as defined by AISC 341 Chapter E parts E2 and E3, the connections to be used are subject to a “conformance demonstration” which can involve specific laboratory testing or the use of “pre-qualified connections”, as approved by AISC 358. There are a number of public domain connection solutions for moment-resisting frames under high seismic loading available as prequalified designs in AISC 358, however none of these pertain directly to hollow structural section members. There are, however, two proprietary seismic moment connection types at present that do have prequalification status, for use with HSS columns.  They are Sideplate (www.sideplate.com) and ConXtech (www.conxtech.com). Please refer to their websites and AISC 358 for additional information.

Question: I have a mitered tube steel truss connection (round sections) with a small angle between the chord and the branch, 16 degrees. The chord is a 10 inch tube and the branch is a 6 inch tube. AISC Design Guide 24 limits you to a 30 degree minimum branch angle. Is that one of the limitations of the underlying research, or is there a different reason for the limitation? Since the capacity has the sinϴ of the branch angle in the denominator, I would guess that using a 30 degree angle in lieu of the 16 degree angle should provide a conservative result, is that correct? I meet all of the other limitations of AISC 360-10 Table K2.1A. The angle I’m using is a limitation of the structure’s geometry. It’s a submerged structure, so I’d like to use a mitered joint rather than a knife plate that would leave the interior of the tube open to the water.

Secondly, AWS D1.1 Figure 3.5 (AISC 13th Edition Manual Table 8-2) limits you to a 30 degree minimum heel angle (ϒ). I understand that I can’t have a pre-qualified joint in the small portion of the connection that is less than 30 degrees, but I’m planning on neglecting that portion of the weld for capacity and only using the weld there to seal the connection. Is there anything else to watch out for in the weld at small acute angles such as this?

Answer: The limitation requiring acute branch angles greater than 30 degrees, in AISC Design Guide No. 24 and AISC 360 (which is taken from International Institute of Welding recommendations), is primarily due to the problems of producing good welds in the heel positions at very low angles. Smaller angles may be acceptable, although not pre-qualified, if you can confirm with the fabricator that proper structural quality welds can be achieved. The difficulty is greatest for square/rectangular HSS connections, but tends to be less of a problem with round-to-round HSS connections – which you have – because that small acute angle is only at one point (the chord crown position). At other positions around the round branch member this local intersection angle “opens up”. I would agree that a fully-sealed welded joint is preferable to a knife plate connection, in your situation. Thus, in conclusion, you need to liaise with the fabricator to obtain confirmation that a specific effective throat can be obtained in the heel, while welding all around.

Question: I am looking for some references regarding the use of through bolts in HSS connections. Specifically, guidance on whether or not the through bolts need to be designed for bending.

Answer: The main concern with “through-bolts” is that when they are tightened the walls of the HSS are deformed and pulled towards each other (inwards). To get over this problem a tube or pipe is often inserted inside the HSS, and the bolt is passed through this pipe. When the bolt is then pre-tensioned the HSS walls remain in alignment and the pipe is placed in compression. With this arrangement there should not be a design concern regarding bolt bending.

Question: The HSS storage racks (shown in the photos below) have been left outside and exposed to the weather for the last few years, with the columns open at the top. The racks have a good bit of rust on a large majority of the members. The posts have major cracks at the welded seams. No analysis has been performed to determine the actual cause of failure, but it is our opinion that water was trapped in the tubes and the pressure generated by freeze-thaw cycles caused the posts to crack at the welded seams. Based upon the locations of the cracks observed in the field, the water levels could have been up to six feet high. The posts vary in thickness (1/4” & 3/8”), but the cracks occur in multiple members. The major cracks occur in the bottom section. We have been asked to provide a repair detail for the steel. We are still determining the best method; currently we are looking at grinding out the existing weld, pulling back the tube with clamps, and re-welding the tubes. Once this is done we will also propose to have a plate welded to the face of the split side. Any thoughts or recommendations would be greatly appreciated.

Answer: Judging by the photographs and your description, it does sound like the damage has been produced by freezing of water within the hollow sections. This is a good reminder that HSS, if exposed to the elements, should be properly sealed or provided with a reliable method for drainage of collected internal water. Splitting of the corners of square and rectangular HSS, due to internal ice build-up, has also been experienced before. Before performing repairs ensure that all water is removed from the HSS. This can be achieved by drilling small holes in the HSS at the base. For repair, I do not think that you need to replace the HSS longitudinal seam welds AND weld on cover plates to the split HSS face. Just the latter should be sufficient (and easier to do) if the HSS rack is now going to be used indoors. The cover plates can be less than the HSS width and fillet-welded (even intermittently) to the HSS flat. These plates will ensure that the member behaves as a closed section under compression loading, and the addition of more steel will moreover locally enhance the cross-sectional properties.

Question: Our company is designing a steel truss with wide flange chords and square/rectangular web members. I see here in your book, “Hollow Structural Section Connections and Trusses” that rectangular webs are only permitted in overlapped K-connections. I am curious as to why they are not allowed in gap connections. Is this a hard rule, or is there simply a reduction in capacity if the rectangular webs are used in gap connections? Where would you recommend I get more information on the subject?

Answer: As you noted, the book “Hollow Structural Section Connections and Trusses – A Design Guide”, 2nd. edition, by J.A. Packer and J.E. Henderson, Canadian Institute of Steel Construction, 1997, does include connection design provisions for HSS web members welded to wide flange (I-shaped) chord members. In Table 3.4(a), it does indicate that the HSS web members are restricted to aspect ratios of 1.0 for K and N gapped connections, but 2.0 for K and N overlapped connections, when using I-section (wide flange) chord members. Research on HSS web member-to-I section chord member connections is actually very limited. In the latest international guidelines on this topic [1] – which are now a draft international standard (ISO 14346) – the range of validity for square/rectangular HSS web members to I-section chord members has now be restricted to HSS aspect ratios of 1.0 for both gapped and overlapped K and N connections, to reflect the scope of the supporting research. References: [1] IIW, 2012. “Static design procedure for welded hollow section joints – Recommendations”, IIW Doc. XV-1402-12, International Institute of Welding, Paris, France.

Question: I have a Warren truss with round HSS branches welded to square HSS chord members. This appears to be outside the scope of the AISC 360 Specification, so is there any way that I can check the adequacy of these connections?

Answer: Yes. You are correct that AISC 360 Chapter K does not cover such connections, which is also confirmed in the Commentary to this chapter, but it also states that for … “round branch members joined to a square or rectangular chord member … other verified design guidance or tests can be used”. Fortunately, some experimentation on such K/N gapped and overlapped connections has been performed (see the test specimen photographs below) and research (Packer et al., 2007) has shown that the round branches can be successfully replaced (for calculation purposes) by square branch members.

The failure modes (limit states) of statically loaded hollow section connections are predominantly related to the branch footprint on the chord (hence branch perimeter) or branch cross-sectional area. Both the perimeter and area of a round and square HSS, having the same width or diameter (i.e. B=D) and being relatively thin, are approximately in the ratio of π/4 = 0.785. It has hence been shown that a “conversion method” can be implemented whereby round branches, of diameter D_b, are replaced by equivalent square branches, of width B_b = (π/4)D_b and the same thickness, and then design rules for rectangular HSS truss connections (i.e. AISC 360-10 Table K2.2) can be used.

Reference:

Packer, J.A., Mashiri, F.R., Zhao, X.L. and Willibald, S., “Static and Fatigue Design of CHS-to-RHS Welded Connections using a Branch Conversion Method”, Journal of Constructional Steel Research, Vol. 63, No.1, 2007, pp. 82-95.