METHOD FOR REDUCING THE COST OF CORROSION TESTING OF REINFORCING STEEL

AN ANALYTICAL PROCEDURE WAS DEVELOPED THAT WILL PERMIT A TEST TO BE CURTAILED AFTER A GIVEN PERCENTAGE OF THE SAMPLES HAVE BEEN TESTED OR WHEN THE TEST IS STOPPED AFTER A SPECIFIED NUMBER OF DAYS. THE ANALYSIS THAT WAS MATHEMATICALLY DERIVED WAS APPLIED TO EIGHT SETS OF DATA SUPPLIED BY THE MATERIALS AND RESEARCH DEPARTMENT. THE ANALYSIS PROVIDED FOR MAXIMUM LIKELIHOOD ESTIMATES OF THE MEAN AND STANDARD DEVIATION FROM A CENSORED NORMAL POPULATION AS WELL AS ESTIMATES OF THE VARIANCE OF THE MEAN AND STANDARD DEVIATION. THESE ESTIMATES PERMITTED COMPARISON OF THE MAXIMUM ERROR EXPECTED FROM A 100 PERCENT SAMPLE TO THE MAXIMUM ERROR EXPECTED FROM REDUCED SAMPLING OF 75, 50 AND 25 PERCENT. IN ADDITION, THE NUMBER OF SPECIMENS REQUIRED TO COMPENSATE FOR THE REDUCED SAMPLES WAS DETERMINED. IT WAS FOUND IN APPLYING THE METHOD TO THE EIGHT SETS OF DATA THAT EXCELLENT CORRELATION WAS OBTAINED, ESPECIALLY AT LEVELS GREATER THAN 50 PERCENT OF THE SAMPLE. A BIMODAL NORMAL DISTRIBUTION WAS DETECTED IN THE DATA AND A CORRECTION WAS DEVELOPED AND MAY BE APPLIED IF SUCH A CORRECTION IS DESIRED. HOWEVER, TYPICAL TESTING VARIATION HAS THE SAME MAGNITUDE AS THE CORRECTION, MAKING THE CORRECTION UNNECESSARY. /AUTHOR/

  • Corporate Authors:

    Department of Transportation

    1200 New Jersey Avenue, SE
    Washington, DC  United States  20590

    California Division of Highways

    Bridge Department
    Sacramento, CA  United States  05914

    Bureau of Public Roads /US

    400 7th Street, SW
    Washington, DC  United States  20590
  • Authors:
    • Benenson, H
  • Publication Date: 0

Subject/Index Terms

Filing Info

  • Accession Number: 00212070
  • Record Type: Publication
  • Source Agency: Bureau of Public Roads /US
  • Report/Paper Numbers: D-3-11
  • Files: TRIS, USDOT
  • Created Date: May 13 1970 12:00AM