State highway transportation organizations should be encouraged to furnish data on painting of weathering steel bridges to the FHWA and SSPC. The type of data that might be delivered could include:
• Costs for spot repairs of coated weathering steel and painting of entire structures.
• Type of surface preparation employed, including details of abrasive used, finish achieved prior to application, e.g., SSPC-SP 10 Near White Metal Blast Cleaning, and surface profile achieved.
• Generic description of coating system, including dry film thicknesses of coating system components, and method of application.
• Condition of steel before application, i.e., degree of rusting or pitting, extent of salt contamination and deicing salt use, temperature and relative humidity at time of application.
• Performance history of installed system based on regular inspections of painted structure. Extent of rusting or degree of undercutting would be useful information, which could be reported either based on standard American Society for Testing and Materials (ASTM) practices such as ASTM 0 610 for rusting, or on the basis of percent area affected. Photographs showing structures inspected would be invaluable. A standard form for recording the data could readily be prepared based on existing SSPC or ASTM documents.
To improve the correspondence between short-term laboratory screening tests and eventual field performance the following are suggested:
• The testing regimen should reflect eventual service requirements. This may include knowledge of such factors as time of wetness, temperature extremes, exposure to sunlight, rainfall and typical components of rust product. This effectively precludes dependence on salt fog testing in accordance with ASTM B 117.
• Laboratory tests should incorporate cyclic elements such as periods of drying, exposure to ultra violet radiation, and the use of a corrosive salt mixture close to that found in service. Replicate laboratory screening tests and field evaluations, using the same combinations of specimens and coating samples in both. This should improve comparisons made between the two and could potentially shorten the overall testing cycle. Preliminary qualification could be made based on an improved short term laboratory testing regimen, final qualification coming from the field exposure results.
• Require larger numbers of samples be used in either field or laboratory evaluations, this permits earlier determination of coating failures.
NOTE: If five or fewer repeat specimens per coating are used, field evaluations demand longer periods of exposure, often up to 10 years.
• Demand the use of statistical evaluation of results for either form of coating evaluation program. Appropriate tests include correlation analysis and ANOVA
• Make this type of experimental design requirement a part of future research proposal requests, i.e., spell out how long you expect the field exposures to last or what degree of statistical probability you would like for the derived answers. Tests exist by which one can determine the optimum sample size for a specific set of statistical requirements.