Bridges, buildings, and other structures commonly include designed bolted connections of steel beams, girders, and other structural members using connection/splice plates of various sizes and configurations. High-strength bolts are used to secure the connections. The holes in the steel members and the connection/splice plates are larger (typically 1/16–1/8 of an inch larger) than the bolt shafts to enable the bolts to be inserted and tensioned against the washers and nuts. The interface of the connection/splice and the structural member is called a faying surface. Faying surfaces are not required to be coated, but are often protected to prevent corrosion at the interface and rust bleed on coated surfaces adjacent to the connection. When a designer elects to coat the faying surfaces, the coating used on these surfaces must have slip-resistant properties to reduce fatigue on the connection during loading/unloading cycles (e.g., vehicles traveling across a bridge deck). Therefore, before use, the coatings specified for these faying surfaces must be tested and classified for slip coefficient properties. Once the classified coating is applied, the connection points are masked to prevent subsequent coats from inadvertently contacting these areas.
There is no stipulation for specific coating types that must be used in these bolted connections, although zinc-rich primers are common and typically have slip resistance properties. Other generic coating types, including polyamide epoxy primers, have been tested and used. Other products, such as thermal spray coatings (various alloys, with and/or without a sealer) and roughened hot dip galvanized steel may be viable candidates, provided they have been tested and demonstrate slip-resistant properties.
The American Institute of Steel Construction (AISC) and the Research Council on Structural Connections (RCSC) publish the Specification for Structural Joints Using High Strength Bolts. The specification comprises ten sections, which only minimally address testing of coatings used in bolted joints. However, Appendix A, “Testing Method to Determine the Slip Coefficient for Coatings Used in Bolted Joints” contains four sections relating to coating testing (General Provisions; Test Plates and Coating of the Specimens; Slip Tests; and Tension Creep Tests). The focus of this article is on the testing required in Appendix A.
Section A1.2 of Appendix A defines Essential Variables, which if changed, will require retesting of the coating to determine its mean slip coefficient. The Essential Variables are not dictated by the specification; rather, they are to be established by the coating manufacturer before testing. There are four Essential Variables, including the cure time (the time interval between coating application and testing), which establishes the minimum curing time prior to field assembly of the joint; any special curing procedures (when they are different from the Product Data Sheet); maximum coating thickness as measured according to SSPC-PA 2, Procedure for Determining Conformance to Dry Coating Thickness Requirements; and the composition of the coating, including the method of manufacture and the type and amount of thinner (if any) to be used.
Appendix A does not provide a tolerance for coating thickness but states only that 2 mils must be added to the coating manufacturer’s maximum thickness (to ensure that a casual buildup of the coating due to over-spray and other causes does not jeopardize the coating’s performance). If the maximum thickness is 4 mils, the test thickness is 6 mils. However, applying exactly 4 mils of coating to a roughened surface in the shop/field (or 6 mils in the laboratory) is essentially impossible. Therefore, a reasonable thickness tolerance should be established before testing. For example, Section 9.1 of SSPC-PA 2 (2012) states, “A minimum and a maximum thickness are normally specified for each layer of coating. If a single thickness value is specified and the coating manufacturer does not provide a recommended range of thickness, then the minimum and maximum thickness for each coating layer shall be +/- 20% of the stated value.” In this case, the tolerance on 6 mils would be 4.8–7.2 mils. If +/- 20% is too wide of a range, then +/- 1 mil may be a reasonable compromise. The test report should indicate the average thickness of the coating applied to each test specimen.
While the method of application is not listed as an Essential Variable, Section A2.2 (Specimen Coating) states that the coatings are to be applied to the test specimens by the same method to be used on the structure itself.
Correction of Coating Thickness Deficiencies
When coatings are applied in the shop or field and the measured dry film thickness is less than specified, it is a common practice to apply a build-up coat (within the manufacturer’s recoat interval) to achieve the specified film thickness. Conversely, if a coating is applied too thick, sanding or screening is used to reduce the applied thickness, or the thickness is not corrected and the excessive thickness is accepted by the facility owner after coating manufacturer’s approval. These corrective practices (sanding/screening) alter the surface of the coating or present a potentially weakened interface (build-up coat) and should not be performed on faying surfaces either in the laboratory for testing or in the shop/field during actual application. Therefore, if thickness deficiencies are found in the faying surface areas, the coating should be removed and reapplied to the correct thickness. The RCSC Standard does not currently address this potential issue.
Test Plate Design and Surface Preparation
There are two test plate designs described in Appendix A, including a Slip Test Plate and a Tension Creep Plate.
The Slip Test Plate is fabricated from steel with a yield strength of 36–50 ksi (or 36,000–50,000 psi). The plates measure 4 x 4 x 5/8-in. and contain a 1-in. (± 1/16-in.) diameter hole centered 11/2-in. from one edge. The top and bottom edges must be milled flat, and surfaces of the plates must be flat enough to ensure reasonably full contact over the faying surface. No burrs or other defects are permitted (Figs. 1 and 2).
The Tension Creep Test Plate is fabricated from the same type and strength steel; however, the plates are larger (4 x 7 x 5/8-in.) and contain two 1 in. (± 1/16-in.) diameter holes centered 11/2-in. from both 4-in. edges. The top and bottom edges are not required to be milled flat; however, the surfaces of the plates must be flat enough to ensure reasonably full contact over the faying surface. As was the case with the Slip Test Plates, no burrs or other defects are permitted (Fig. 3).
Before mechanical methods of surface preparation, the surfaces are solvent cleaned in accordance with SSPC-SP 1 to remove grease, oil, or other fabrication lubricants. The faying surfaces are subsequently blast cleaned to the required surface profile depth (e.g., 2–3.5 mils). Details related to surface preparation are undefined by the RCSC specification but must be reported by the testing laboratory. The edges of the test plates and the holes are not prepared by mechanical means, nor are they coated.
Test Plate Mounting, Coating Application, and Curing Procedures
The procedures for mounting and securing the test plates are not prescribed by the RCSC specification. However, the test surfaces must be free of surface defects so handling the freshly coated plates is problematic. Therefore special racks or trays can be used to hold the test plates in place horizontally (using compression fit along the non-milled edges) so that the faces can be sprayed and cured horizontally (and without handling), in order to prevent runs, sags and defects in the applied film (Figs. 4 and 5). Applying the coatings to test plates mounted horizontally is especially important because the specification requires the addition of 2 dry mils to the coating manufacturer’s maximum thickness (which may exceed the sag resistance thickness for a given coating). Fifteen slip test plates and nine tension creep test plates are required for testing.
Coating materials are applied (Fig. 6) and cured according to the manufacturer’s written instructions, adding thinner if required by the manufacturer. The type and amount of thinner added is an Essential Variable and must be reported on the Certificate of Testing (described later). The batch/lot numbers of the coating components and thinner are also recorded and listed on the test certificate. An owner/specifier may require specific curing conditions (air temperature, humidity, and time). Conditions should be monitored throughout the curing period using a recording hygrothermograph or a digital psychrometer.
Coating Thickness Measurement and Selection of Contact Surfaces
Coating thickness is measured according to SSPC-PA 2. The average of the measurements for each contact surface is recorded and reported. Contact test surfaces with conforming coating thickness values (and with similar average thickness values) are selected.
A Slip Coefficient test assembly consists of three test plates: a center plate with two contact surfaces and two outside test plates with one test surface each. Five replicate assemblies must be tested. A Tension Creep test assembly is similar, but only three replicate assemblies must be tested.
Slip Coefficient Test Procedure
Once the mating surfaces are selected (based on similar average coating thickness values), the test plates are loaded onto a 7/8-in. threaded rod located in the center of a horizontal load cell (Fig. 7). The center test plate is inverted 180 degrees as it is loaded onto the rod so that the edge of the plate is higher than the two end plates. This is the plate that receives the vertical load during testing. A clamping force of 49±0.5 kips (49,000±500 lbf, or pounds of force) is applied to the test assembly using a horizontal calibrated ram operated using hydraulic pressure, or a ram in conjunction with a load cell. This load is maintained throughout the testing process and is monitored using a digital readout (Fig. 8).
Once the clamping force is applied, the vertical load cell platen is lowered so that it contacts the top (milled) edge of the center test plate. After 1 kip (1,000 lbf) of load is applied to the vertical platen, slip deflection monitoring gages are attached and engaged. The vertical compression load is applied at a rate less than or equal to 25 kips/minute (or a maximum of 0.003 in. of slip displacement per minute). Each assembly takes 10–15 minutes to mount and test.
The test is terminated when 0.05 in. of slip (or greater) occurs. The slip displacement is displayed digitally and is monitored and recorded on an X-Y plotter. Typical slip responses are illustrated in the RCSC specification (Fig. 9).
Slip Coefficient Calculation
The mean slip coefficient (ks) is calculated as
Sample data for the slip coefficient calcuation is shown in Tables 1 and 2.
Tension Creep Test Procedure
The second phase of the testing, the tension creep test, is undertaken, provided the slip coefficient testing produces acceptable results. This test is longer (1,000 hours or six weeks) than the slip coefficient test, which takes only a couple of hours. For this phase, a “chain” of nine 4 x 7-in. test plates is assembled using A490 bolts (Fig. 10). Once assembled, the chain is suspended from the tension creep frame, and the vertical load is applied and monitored using a load cell. The applied load is based on the specified minimum bolt pretension and the slip coefficient class established by testing (A, B or C). For example, 25.9 kips load is applied for Class A slip coefficient using A490 bolts, while 39.2 kips load is applied for Class B slip coefficient using A490 bolts. Once the prescribed tension is applied, the chain of test plates is locked in place using a large nut. This tension is maintained throughout the test period.
Movement (creep) is monitored by micrometers, which are set to “0” within 30 minutes of applying the load. The load is maintained for 42 days (1,000 hours). Any displacement (revealed by movement from the “0” position on the micrometers) is recorded.
Three replicate assemblies are tested concurrently. No single assembly can exceed 0.005 in. of displacement. After 1,000 hours, the load is again increased to the design slip multiplied by 2x the average clamping force. Once this final load is applied, the average creep displacement cannot exceed 0.015 in. for all three assemblies. Table 3 shows sample data for tension creep testing.
Once the slip coefficient and tension creep testing is complete (and the coating passes both tests), the coating is classified. According to the RCSC specification, the mean slip coefficient (μ) can be categorized as Class A, B, or C. A “Class A” slip coefficient is 0.33 minimum (uncoated, clean mill scale or coatings on abrasive blast cleaned steel); a “Class B” slip coefficient is 0.50 minimum (uncoated, abrasive blast cleaned steel or coatings on abrasive blast cleaned steel); and a “Class C” slip coefficient is 0.35 minimum (roughened, hot-dip galvanized surfaces). Note that ANSI/AISC Specification 360-10, Specification for Structural Steel Buildings, lists 0.30 slip coefficient for Class A as opposed to 0.33.
A Certificate of Testing accompanies the test report. The certificate lists
The test certificate is very important because it lists the Essential Variables under which the product was tested and classified. The certificate should be a required submittal from the coating manufacturer that has been selected to provide coatings for a project. Further, the product should be applied to the faying surfaces in the shop/field in conformance to the listed Essential Variables.
Note that surface preparation (cleanliness and profile depth/shape) is not listed on the test certificate (they are not considered Essential Variables by the RCSC specification), and the minimum dry film thickness is not listed (only the maximum). The minimum may be established by the coating manufacturer on the product data sheet.
Although the testing and data management procedures are fairly well defined in Appendix A of the RCSC specification, there is a need for coatings industry research relating to the test plate preparation and coating procedures. The potential research initiatives include:
1. What is the effect (if any) of surface profile shape on the slip coefficient properties of coatings?
The shape of the surface profile generated by abrasive blast cleaning operations can be angular or rounded, depending on the shape of the abrasive used. Steel fabrication shops typically use steel shot or a blend of steel grit and shot. Field contractors typically use angular abrasives (mineral/slag or recyclable steel grit). The resulting shape of the surface profile influences the surface area of the steel. (Angular abrasives produce a denser peak/valley pattern than shot abrasive, which results in increased surface area.) It is unknown whether the texture of the steel surface influences the slip coefficient properties of the applied coating. Surface profile shape is not currently considered an Essential Variable in the RCSC standard. Note that AASHTO R31, Evaluation of Protective Coating Systems For Structural Steel, requires the use of 100% steel shot (S280) for the test “selected to create a worst-case scenario,” which suggests that the surface texture may influence the slip coefficient properties.
2. What is the effect (if any) of surface profile depth on the slip coefficient properties of coatings?
The depth of the surface profile can range from <1 mil to five to six mils, depending on the abrasive size, hardness of the steel surface, and other influences. The surface profile depth is typically based on the thickness of the coating to be applied and is typically expressed as a range. While some specifications may invoke a one- to two-mil surface profile, others may invoke a three- to four-mil surface profile. It is unknown whether the depth of the surface profile influences the slip coefficient properties of the applied coating. Surface profile depth is not currently considered an Essential Variable in the RCSC standard.
3. Is there a difference in slip coefficient properties when a coating is tested over a surface that has been power tool cleaned (i.e., SSPC-SP 11 or SP 15) versus abrasive blast cleaned?
Maintenance of existing bridge structures may include replacement of connection plates. Power tool cleaning of the bridge beam connection areas is sometimes permitted to eliminate the need to mobilize abrasive blast cleaning equipment and to contain as well as properly manage a waste stream for preparation of relatively small areas. These areas may be prepared to SSPC-SP 11, Power Tool Cleaning to Bare Metal, or SSPC-SP 15, Commercial Grade Power Tool Cleaning, which allows up to 33% staining on the prepared surfaces. Both standards invoke a minimum one-mil surface profile; they do not invoke a maximum surface profile depth. The new connection plates coming from a steel fabrication shop are typically abrasive blast cleaned and primed. Accordingly, there is a potential to mate faying surfaces having two different surface textures (surface profile depth and shape). It is unknown whether these differences influence the slip coefficient properties of the applied coating. The method of surface preparation and the degree of cleanliness are not currently considered Essential Variables in the RCSC standard.
4. Is there a curing “window” (both a minimum and a maximum set time prior to bolt-up)?
The minimum cure time, established by the coating manufacturer, is considered an Essential Variable and is listed on the Test Certificate. It is not dictated by the Standard. Organic zinc-rich primers contain an organic binder (most commonly epoxy or urethane), which can become harder over time. It is unknown whether an extended cure time prior to bolt-up (e.g., one month) influences the slip coefficient properties of the bolted connection.
5. What is the effect of using a different type of thinner (acceptable for use by the coating manufacturer) on the slip coefficient properties?
The type of thinner used to reduce the coating is considered an Essential Variable and is listed on the Test Certificate. However coating manufacturers often list more than one type of thinner that is compatible with the coating. The reasons for listing multiple thinners vary, but generally are due to different application conditions (“normal” versus “hot/windy”) and environmental regulations (thinners containing exempt solvents for use in certain areas of the country). Does a coating manufacturer need multiple Test Certificates for a single coating to cover all acceptable thinners, so that the appropriate thinner can be used at the time of application? That is, it is unknown whether the type of thinner used for reduction has an effect on the performance properties of the coating (given that the thinner is compatible with the coating).
6. What is the effect of using lesser or greater amounts of thinner?
The amount of thinner used to reduce the coating is considered an Essential Variable and is listed on the Test Certificate. Thinner amounts are established by the coating manufacturer. Coating manufacturers often list a maximum amount of thinner that is acceptable for use (e.g., up to 10%) and the amount may vary depending on application method and prevailing conditions of temperature, humidity and wind. Does a coating manufacturer need multiple Test Certificates for a single coating to cover various amounts of thinner (up to the maximum allowable amount), so that the appropriate amount thinner can be used at the time of application? That is, if “up to 10%” thinner (reducer) is permitted by the coating manufacturer and the coating is tested using 5% reducer, but 10% is used in the shop or field due to prevailing conditions, it is unknown whether the slip coefficient properties of the coating are affected by the variation in reducer amounts.
7. Five replicate assemblies are tested for slip coefficient and then averaged to generate a single slip coefficient value. Is there an acceptable standard deviation between the replicate trials? That is, what determines an “outlier”?
Occasionally, one of five replicate tests generates a slip coefficient value that is considerably different than the other four. However, the RCSC standard does not address standard deviation between test results, only that the five replicates are averaged. Accordingly, a low slip coefficient value (on one of the five replicates) could result in a Class B rating dropping to a Class A rating, or even non-classification. ASTM Standard Test Methods (methods that produce a value) are required to have precision and bias statements prepared based on interlaboratory studies that generate repeatability and reproducibility data. ASTM Subcommittee D01.46, Industrial Protective Coatings, has expressed interest in creating a Work Item and subsequently drafting and balloting a test method for slip coefficient and tension creep testing, which would address the repeatability concern. However, there are a very limited number of laboratories that are equipped to perform this test, so conducting a statistically valid study is challenging.
8. Mating of Dissimilar Coatings
Traditional slip coefficient testing mates the same coating on the faying surfaces. In some cases, there may be a need to obtain Class A or B slip coefficient certifications for an inorganic zinc primer (applied to new connection/splice plates in a shop and shipped to the field) to a field-applied organic zinc primer (on existing steel during maintenance of a bridge structure) that may have different thickness requirements. A project engineer should realize that coating manufacturers may not have this test data on hand, and that testing and certification can take two or more months to generate. Coordination with the shop (requiring them to apply the same product as the field contractor will use) will alleviate this issue, provided coating thickness is addressed.
About the Authors
The Journal Of Protective Coatings & Linings ©2014 Technology Publishing Company
like this article? you might also enjoy…
2 thoughts on “Understanding Slip Coefficient and Tension Creep Testing of Coatings Used in Slip-Critical Bolted Connections”
Very interesting study. I admire your vision and leadership in the field of protective coatings. KTA-Tator has a strong standing in NY Metro area. I work for the Port Authority for 25 years and remember your company since the early 90’s.
Your study has an undertone that drives into the science a bit to deep. Indeed it’s hard to use statistics…
I suggest to consider the mindset of code writing folks . After all, codes are written to reach a consent between science and pragmatic approach for everyday construction.
I suggest to concentrate on new quick tests that take into consideration dynamic loads. E.g. cyclical with high frequency with loads modulation, that could be co-related to static field tests by ICC special inspectors. Statistic co-relation vs. targeted application present a challenge, though. Yet, it’s doable .
This is an excellent article and you have brought up some of the same questions that I have regarding this testing particularly the impact on slip due the nature and depth and type of surface profile, dissimilar coatings and surface prep, curing time and thinning. Now the hard part, answers to them. Possible research by FHWA or NTPEP?
Comments are closed.