Guidelines for Concrete Mixtures Containing Supplementary Cementitious Materials to Enhance Durability of Bridge Decks (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

111 Introduction In Step 5, the BTC and the BPC were identified. The BTC is the produced concrete that performed better than the other produced concretes tested. The BPC is a concrete identified based on numerical modeling as having performance that exceeds that of the BTC. In Step 6, additional testing and subsequent analysis will be performed to confirm that the BTC and BPC actually perform as suggested by the modeling performed in Step 5. The objective of this step is to select the BC. The Step 6 testing and analysis are intended to provide confirmation that the performance of this mixture is better than the BTC as pre- dicted so that this selection can be made with a high level of confidence. In some cases, time or budget constraints will not allow for additional testing. In these cases, the BTC should almost certainly be selected as the BC and recommended for use. Because no testing has been done on the BPC, it is highly risky to select the BPC as the BC without actually mixing a batch of the BPC and subjecting it to the most critical performance tests for the application. The user will have to consider experience or well-documented research in making this decision. Confirmation Testing The following questions need to be answered regarding confirmation testing: • What concrete mixtures should be mixed and tested? • What performance tests should be run? • How should the test data be analyzed? • Did the data and models in Step 5 accurately predict the performance of the BTC and BPC? • What concrete mixture should be selected as the BC and recommended for the application? Mixtures and Tests If possible, batches of both the BTC and the BPC should be mixed and should be tested using all the performance tests that were used in the analysis in Step 5. However, in some cases, conducting all performance tests for both the BPC and the BTC will not be feasible. In these cases, the first consider- ation should be whether the BPC appears to provide a sub- stantial improvement over the BTC (e.g., as indicated from the predicted overall desirability for the BPC and the BTC). However, a more detailed examination/comparison of the predicted test results for the two concrete mixtures should be done to determine if the BPC should be pursued. If the pre- dicted performance of the BPC is not significantly better than the performance of the BTC, then there is little reason to test the BPC. The confirmation test budget could be used to run more testing on the BTC to confirm that the high perfor- mance observed in the earlier tests can be repeated. If the predicted performance of the BPC is substantially higher than the BTC for at least some important performance tests, as many performance tests as practical should be run on new batches of the BTC and the BPC. If not all tests can be performed, the number of performance tests should be reduced: • Any performance test that is absolutely critical to the ap- plication should not be eliminated from the confirmation testing. Performance tests that are very expensive or take an extremely long time are probably the most likely tests to be eliminated. However, because the BPC has never been tested, eliminating a test that is critical to the application will result in a deficient evaluation of the BPC, making it unsuited for selection as the BC. • A side-by-side comparison of the predicted performance of the BPC and the BTC should be observed. In many tests, the predicted performance of the BPC may not differ sig- nificantly from that of the BTC. This difference should be S T E P 6 Perform Confirmation Testing and Select Best Concrete

evaluated in terms of the individual desirability and not the actual response value, because the individual desirability will influence overall desirability. Performance tests that produced little difference are not likely to yield much new information in the confirmation testing and thus should be the first to be eliminated if budget and time constraints require it. If all mixtures performed well in freezing and thawing resistance tests as a result of good air void systems such that the predicted desirabilities are all essentially 1, then repeating this test is not likely to contribute much in- formation and the test may be eliminated. However, the air content of the hardened concrete could still be measured as an indication of good cyclic freezing resistance. Based on the above guidelines and the judgment of the users, the concrete mixtures (BTC and/or the BPC) and a set of tests can now be selected for determining the overall desir- ability from the confirmation testing. Data Analysis: Model Checking Once the confirmation testing is completed, the data should be compared with the results of the main test pro- gram. These data may be added to the plots already created. Although not as important as determining the BC, check- ing the accuracy of the model prediction from Step 5 is a use- ful step for gaining confidence in the use of the regression models. The accuracy of the model prediction can be evalu- ated for all the performance tests that were repeated in the confirmation testing. If only a subset of the performance tests is run in the confirmation testing, a predicted overall desir- ability for the BTC and the BPC for only those tests should be calculated from the regression models used in Step 5 (this predicted overall desirability will be somewhat different from the one used to select the BTC and the BPC). Next, an overall desirability from the actual confirmation test results needs to be calculated for both the BTC and the BPC. Comparing the predicted overall desirability and the actual overall desirability is a useful first step, but the user should also compare the individual predictions of each test result with the actual results obtained in the confirmation testing. In some cases, calculating a percentage difference is useful, but judgment is needed to determine what consti- tutes poor prediction accuracy. If the models predict well, then more confidence can be placed in the prediction and trends observed in the previous step. It also builds confi- dence in the predicted values for the BPC for any tests that were not performed in the confirmation testing. If there are major problems with the prediction accuracy, then all deci- sions should be based only on actual data, which in most cases would favor the BTC that has been batched and tested more often than the BPC. Final Selection of the BC It is expected that the concrete mixture that has the high- est overall desirability will be chosen as the BC, although ad- ditional considerations may affect this selection. The most important consideration is that the results for each of the im- portant performance tests should indicate that the concrete mixture will perform well on those performance characteris- tics that are critical to the application. If there is little differ- ence between the overall desirabilities of the BTC and the BPC, then individual response performance and desirabilities should be examined, and the mixture that performs best in the tests that are most significant for the application should be chosen. The cost, availability of materials, and ease of pro- duction may also be considered. If the differences are not suf- ficient to justify the production of the mixture with the higher desirability, then that mixture should not be chosen. Finally, the accuracy of the test program conclusions should be con- sidered. If the overall desirabilities of the BTC batches in the original test program and the confirmation tests are different by a greater amount than the BTC and BPC, the difference is likely within the range of repeatability of the experiment. Then, the primary consideration should be individual test performance with secondary consideration placed on cost and other factors. Example from Hypothetical Case Study The BPC and BTC identified in the hypothetical case study were tested according to the revised list of test methods. Table S5.3 lists the responses that were included in the calculation of the overall desirability for the confirmation testing. The confirmation test program was limited to those responses that showed significant performance differences for BTC and BPC and that could be completed in the available timeframe. Therefore, the finishability, modulus of elasticity, and freez- ing and thawing tests were eliminated because the BTC and BPC mixtures were predicted to perform such that a similar desirability would be assigned to that response. The cracking tendency test was eliminated because the test could not be completed in the allotted time. The method used to evaluate the chloride penetration was changed to ASTM C 1556 tested at 56 days because of time constraints. However, because both of the chloride penetration test methods used measure similar performance, the initial and confirmation test pro- grams were considered essentially comparable. The overall desirabilities of these mixtures were determined using the individual desirability functions and the results ob- tained during the confirmation testing. These overall desir- abilities are compared with the predicted overall desirabilities in Table S6.1, which also includes the overall desirability from 112

the original BTC batch calculated using only the tests in the confirmation testing program. Slight differences between the overall desirabilities based on the confirmation round of test- ing and those calculated based on the Step 4 round of testing are due to the different set of responses used in the confirma- tion analyses (see Table S5.3). For the hypothetical case study, the actual and predicted performances of the confirmation BTC and BPC agreed very well. The difference between the actual BPC and BTC per- formance is much greater than the difference between the performance of the original and confirmation batches of the BTC. This result indicates that the test program produced re- peatable results and that the increase in desirability measured in the BPC is a significant and measurable improvement. Tables S6.2 and S6.3 present the actual and predicted indi- vidual responses and corresponding desirabilities for the con- firmation testing for the BTC and BPC that could be used to evaluate the accuracy of the predictions in terms of the test re- sponses and the corresponding desirabilities, one response at a time. The mixture responses that were least well-predicted (i.e., mixtures that showed the greatest percentage difference) in terms of the test result for the BTC and BPC were the elec- trical conductivity and mass-loss scaling tests. However, the corresponding desirability values varied only slightly because the desirability functions placed only limited significance on these differences. In fact, only one desirability prediction was different by more than 5%: the 7-day strength prediction for the BTC differed by 5.2%. The confirmation test results evaluated in terms of the over- all desirability and the good agreement between the test responses and the model predictions used to select the BPC confirm the accuracy of the statistical analysis. This result justi- fies the selection of the BPC as the recommended BC mixture— the optimum mixture based on the available raw materials. 113

114 Tables for Step 6 Mixture Actual Overall Desirability Predicted Overall Desirability % Difference BTC Original Batch (Mixture #8) 0.9615 0.9601 0.1 BTC Confirmation Batch 0.9601 0.9601 0.0 BPC Confirmation Batch 0.9724 0.9700 0.2 seitilibariseD laudividnI sesnopseR laudividnI Property Original BTC Batch (Mixture #8) BTC Confirmation Test BTC Prediction BTC % Difference Response BTC Confirmation Test BTC Prediction BTC % Difference Desirability Included in Confirmation Test Analysis Slump (in.) 7.75 6.25 8.05 -22.4% – – – No Slump Loss (in.) 1.75 2.25 1.89 19.3% – – – No Plastic Air (%) 6.10 7.00 6.34 10.4% – – – No Hardened Air (%) 5.70 7.50 6.09 23.1% – – – No Initial Set (h) 5.50 5.08 5.33 -4.8% 1.000 1.000 0.0% Yes Finishability 11.3 No test 11.8 – – – – No Cracking Tendency (wks) 7.0 No test 7.4 – – – – No Heat of Hydration Temperature Rise (°F) 46 46 45 3.1% 0.957 0.959 -0.2% Yes Shrinkage (% ) -0.0441 -0.0452 -0.0445 1.7% 0.974 0.978 -0.4% Yes Specific Surface Area (in.-1) 408 No test 417 – – – – No Compressive Strength at 7 Days (psi) 5705 6020 5367 12.2% 0.948 1.000 -5.2% Yes Compressive Strength at 28 Days (psi) 7888 7970 7194 10.8% 1.000 1.000 0.0% Yes Compressive Strength at 56 Days (psi) 8460 8520 7793 9.3% 0.997 1.000 -0.3% Yes Modulus of Elasticity (x 106 psi) 4.26 No test 4.25 – – – – No Electrical Conductivity (Coulombs) 1136 778 1143 -31.9% 0.961 0.929 3.5% Yes Scaling: Visual 0.0 0.0 0.1 – – – – No Scaling: Mass Loss (g/m2) 86.7 25.0 93.4 -73.3% 0.993 0.972 2.1% Yes Freeze-Thaw Durability Factor (%) 103.8 No test 103.7 – – – – No Chloride Diffusion Coefficient (x 10-12 m2/s) 1.62 1.88 1.95 -3.8% 0.859 0.853 0.7% Yes Table S6.1. Comparison of actual and predicted overall desirabilities from confirmation testing. Table S6.2. Comparison of individual responses and desirabilities for Best Tested Concrete.

115 Individual Responses Individual Desirabilities Property BPCConfirmation Test BPC Prediction BPC% Difference Response BPC Confirmation Test BPC Prediction BPC % Difference Desirability Included in Confirmation Test Analysis oN – – – %2.2 01.7 52.7 ).ni( pmulS Slump Loss (in.) 3.00 2.49 20.4% – – – No Plastic Air (%) 6.7 6.4 4.0% – – – No Hardened Air (%) 6.3 6.7 -6.0% – – – No Initial Set (h) 6.42 5.66 13.5% 1.000 1.000 0.0% Yes Finishability No test 11.4 – – – – No Cracking Tendency (wks) No test 15.7 – – – – No Heat of Hydration Temp. Rise (°F) 44 44 0.4% 0.960 0.960 0.0% Yes Shrinkage (%) -0.0476 -0.0434 9.6% 0.962 0.983 -2.1% Yes Specific Surface Area (in.-1) No test 424 – – – – No Compressive Strength at 7 Days (psi) 5570 5504 1.2% 0.993 1.000 -0.7% Yes Compressive Strength at 28 Days (psi) 7710 7731 -0.3% 1.000 1.000 0.0% Yes Compressive Strength at 56 Days (psi) 8560 8383 2.1% 0.992 1.000 -0.8% Yes Modulus of Elasticity (x 106 psi) No test 4.24 – – – – No Electrical Conductivity (Coulombs) 244 397 -38.5% 0.988 0.980 0.8% Yes Scaling: Visual 0.0 0.3 – – – – No Scaling: Mass Loss (g/m2) 52.8 183.0 -71.2% 0.984 0.945 4.1% Yes Freeze-Thaw Durability Factor (%) No test 104.0 – – – – No Chloride Diffusion Coefficient (x 10-12 m2/s) 1.28 1.38 -6.8% 0.904 0.897 0.8% Yes Table S6.3. Comparison of individual responses and desirabilities for Best Predicted Concrete.