Cement by Heat of Neutralization

Transcription

1 TRANSPORTATION RESEARCH RECORD D~termining Cement Content of Soil Cement by Heat of Neutralization RonERT SA vuzzo A heat of neutralization test method (a modified by the Bureau of Relamation) for detennining ement ontent of freshly mixed soil-ement is di.sussed. One a aiibration urve is establi hed, a ement ontent determination an be made in approximately 15 min, whih in field appliations has been found to be aurate to within ± l perentage point of atual ement. ontent. A sodium aetate- glaial aeti aid buffer solution is u ed 1 initiate an exothermi reation with the alium hydroxide in the soilement test speimen. During onstrution, rest peimens are obtained diretly from the spreader as soil-ement is plaed, and separation of plu No. 4 (4.75-mm) ieve size material i not required. Results from an interlaboratory testing program were analyzed to establish prei ion and bias of the S\tgge ted te L method. In 1988 the Bureau of Relamation was plaing soil-ement for lope stabilization at Jakson Lake Dam, Wyoming. Beause of gradation harateristis of borrow material, a oilement mix design having 5 perent plu No. 4 (4.75-mm) sieve size material was the most eonomial mixture. There was onern that typial titration methods for determining ement ontent of soil-ement for onstrution ontrol would be diffiult to use and too slow to perform. Titration methods typially require sepurating the plu No. 4 (4.75-mm) sieve si.ze material and inlude t11e time-on urning task of sreening and wa hing the soil-ement mix to obtain test speimens of appropriate size and gradation harateri tis. Thi irumstane prompted initiation of a program to develop a test method for determining ement ontent of soil-ement with the following objetives: (a) separation o{ plu No. 4 (4.75- mm) sieve size material not required for obtaining test speimens; (b) ement ontent determinations in a timely manner i.e., within 15 to 2 min () durable apparatus for field use; (d) ease of performing test method by field per onnel; and (e) auray of ± J perentage point of atual ement ontent for oil-ement mjx design having ement ontents of 3 to 15 perent and up to 5 perent plus No. 4 (4.75-mrn) ieve size material. TEST METHOD Qll6B-1978 Test methods used for determining ement ontent of soilement mixtures were reviewed and evaluated against the five requirements. Test Method Ql16B-1978 Cement Content of Cement Treated Materials (Heat of Neutralization) (1), was U.S. Bureau of Relamation, P.O. Box 257, Mail Code D-3735, Denver, Colo the test method that ame losest to satisfying the five requirements. In the test method, an aidi buffer solution of glaial aeti aid and sodium aetate with an initial ph of approximately 2 is used to reat with the alium hydroxide in a soil-ement test speimen. This reation is exothermi and the heat produed is proportional to the quantity of ement present in the test speimen. The nature of thi proportionality is linear, whih makes it onduive to establishing a alibration urve and determining a orresponding line equation for a given material. Soil-ement test speimens of unknown ement ontent an then be tested by obtaining the heat of neutralization, and the ement ontent an be determined from the alibration urve. Heat f neutralization is defined as the temperature inrea e resulting from the exothermi reation in the aidi buffer solution when mixed with a soil-ement test speimen. The Australian test method speified that a 5-kg soil-ement test speimen be mixed with 1 L of buffer olution onsisting of 15 g of sodium aetate, 24 g of glaial aeti aid and distilled water. A alibration urve is prepared using seven different perentage of ement onten'ts that braket the ement ontent peified in the mix design, i.e., if the target figure is 2 perent ement, perentages of ement of.5, 1., 1.5, 2., 2.5, 3. and 3.5 are u. ed. Calibration te t. peim n are prepared at the water ontent to be u ed during soilement plaement. Dupliate heat of neutralization te ts are performed at eah of the seven ement ontents. Temperature inrea e orre ponding to eah of the 14 tests is plotted ver u ement ontent and a alibration tine is establi bed. Sub e quent soil-ement test speimens o.f unknown ement ontents are tested the beat inrease obtained i evaluated using alibration data, and the ement ontent is determined. RECLAMATION PROCEDURAL CHANGES Test Method Qll6B-1978 was originally developed for soilement mix designs having ement ontents of 3.5 perent or less. At ement ontents greater than 3.5 perent, the mixture of 1 L of buffer solution with a 5-kg test speimen gels into a solid mass, preventing proper mixing. Reommendations by the Main Roads Department for ement ontents greater than 3.5 perent were (a) add additional buffer solution, up to 3 L, to the 5-kg soil-ement test speimen; (b) adjust the mass ratio of buffer solution to test speimen from 1 L to 5 kg for ement ontents less than 3.5 perent, to 2.5 L to 4 kg for ement ontents between 6. and 7.5 perent. It was also stated that any required redution of test speimen mass should

2 18 TRANSPORTATION RESEARCH RECORD 1295 be kept to a minimum and under no irum tane sh uld th oi l-ement test speimen be less than 4 kg. After perf rming a eries of tests, the Bureau of Relamation made the following hanges to the riginal Australian heat of netth alization test meth d: (a) redue the number of ement ontents required for obtaining a alibration urve from seven to three; (b) perform three heat of neutralization tests at eah of the three ement ontent values seleted, to onstrut a alibration urve u ing 9 data points, redued from 14; and () use 1.5 L of buffer solution mixed with a 1.5-kg soil-ement test speimen for mix design on isting of up to 15 perent em 111 and 5 perent plus No. 4 (4.75-mm) sieve ize material. INTERLABORATORY TEST PROGRAM After the hange were inorporated into the test method, an interlaboratory test program was performed in aordane with A TM C82-87, Standard Pratie for Condu1ing n11 Jmerlaboralory Test Program 1 Determine the Preision of Test Methods for Conslrution Material. This te t program was performed to validate the proedurnl hange outlined, soliit u er omments on the revised heat of neutralization test proedure, and provide reliable information from whih preision tatements of the type presribed in ASTM C67-88, Standard Pratie for Preparing Prei ion and Bias Statements for Test Methods for Constrution Materials, ould be developed. The interlaboratory test program wa performed on two oil-ement mix design in two separate phases. Pha e I onsisted of testing soi l-ement speimens oi a fine-grained mix design and Phase II used a oarse-grained mix design. Fineand oarse-grained mix de igns were seleted to determine if th hanges made to the test method were effetive both on a fin -grained mix and on one in whih a signifiant perentage of plu No. 4 (4.75-mm) sieve size material wa pre - ent. The test method hange investigated wt:re (a) reduing the number of ement ontent and alibration te t speimens required to onstrut a alibration urve, and (b) reduing the test speimen ize from 5 to 1.5 kg. Testing laboratories from federal, state, and private setors partii.pated in the te t program. Ten testing laboratories partiipated in Pha e I of the test program and eight testing laboratories partiipated in Phase TI. Partiipating laboratorie were provided with a written proedur anrl individual test peimens of appropriate soil ma eah with a orresponding amount of ement. Eah laboratory wa required to thoroughly mix the soil, ement and water, as speified for eah re t speimen before performing the heat of neutralization test. A moisture ontent of 8 perent was peified for eah test speimen. Phase I For the first phase of the test program, eah partiipating laboratory performed a total of 15 ement ontent determinations on oij-ement test speimen. The peimens onsisted of 5 perent minus %-in. and 95 perent minus No. 4 (4.75-mm) sieve ize material. Nine determinations were performed on test speimen of known ement ontents-three at 5, three at 7 and three at 9 perent ement-to develop a alibration urve and determine a alibration line equation. The remaining six ement ontent determination were performed on dupliate te t speimens designated for eah laboratory as Te t Speimens A, B, and C, but of ement ontent unknown to the partiipating laboratorie. Test Speimens A were prepared at 6 perent ement, B at 7 perent ement and C at 8 perent ement. Phase II The purpose of the eond pha e of the interlaboratory test program wa to determine the effet that a ignifiant perentage of plu No. 4 (4.75-mm) sieve ize material would have on the reliability of the revised beat of neutralization proedure in det:rmining ement ontent. Phase II of the test program on i ted of the same number of test peimens at the same ement ontents as in Phase I. T he gradation f Phase IT test speimen was as follow : Material Size 'I to 1 Y2 in. % to 'I in. No. 4to%in. Minus No. 4 Calibration Test Speimen Results Perent Retained Results obtained from Phase I alibration test speimens (5 perent plus No. 4) are shown in Figure 1. Stati tial analysis performed on data from the 9 heat of neutralization test 3 at eah of the three ement ontents, resulted in a orrelation oeffiient of.981 with a orresponding R 2 value of 96 perent. The solid line shown in Figure 1 represents the best-fit line obtained performing a linear regression through all data points. The narrow set of dashed lines indiates the 95 perent onfidene limit within whih the average value of all data points would fall. The wider set of dashed lines indiates the 95 perent onfidene limit within whih any data point would fall. Results obtained from Phase II alibration test speimens (5 perent plus No. 4) are shown in Figure 2. Statistial analysis performed on data from the 72 heat of neutralization tests, 24 at eah of the three ement ontents, resulted in a orrelation oeffiient of.974 with a orresponding R 2 value of 95 perent. The solid, narrow, and wide sets of dashed lines represent the best-fit line, 95 perent onfidene limits for average data points, and 95 perent onfidene limits for any data point, respetively, as desribed for Phase I alibration test speimen results. Tables 1 and 2 present summaries of Phase I and Phase II alibration test speimen results, respetively. Average, maximum, and minimum values of temperature inrease in degrees Celsius, as well as alulated within- and between-laboratory standard deviations, are provided. Test Speimen Results lnterlaboratory test program results for Test Speimens A, B, and C (5 perent plus No. 4) and Test Speimens D, E,

3 Savuzzo 19 at Correlation Coeffiient =.981 A-squared = 96 perent ,, :;::>.~ ~ Q) z iii Q) I ,.\ ~... # Represents 95% onfidene limit within whih any data point would fall. 6 6 B 9 FIGURE 1 Phase I alibration test speimens. Cement Content - % and F (5 perent plus No. 4) are presented in Tables 3 and 4, respetively. Average, maximum, and minimum values of ement ontent, as well as alulated within- and betweenlaboratory standard deviations, are provided. Preision and Bias Statements Results obtained from the interlaboratory test program were used to develop preision and bias statements as presribed in ASTM C67-88 for the heat of neutralization test proedure as revised by the Bureau of Relamation. Table 5 presents a summary of single-operator and multilaboratory preision and bias values obtained for the alibration test speimens. Values shown in Rows 2, 3, and 5 of Table 5 are alulated in aordane with ASTM C67-88 and are in degr ~s Celsius. Table 6 presents a summary of single-operator and multi Iaboratory preision and bias values obtained for Test Speimens A, B, C (5 perent plus No. 4) and Test Speimens D, E, F (5 perent plus No. 4). Values shown in Rows 2, 3, and 5 of Table 6 are alulated in aordane with ASTM C67-88 and are in perentage points of ement ontent. From the single-operator and multilaboratory preision and bias values presented in Tables 5 and 6, preision statements an be written for speimens ontaining from 3 to 15 perent ement and up to 5 perent plus No. 4 (4.75-mm) sieve size material as desribed in the following setion. CALIBRATION TEST SPECIMENS Single-Operator Preision The single-operator standard deviation of a single test result (a test result defined in this proedure as the average of three separate measurements) has been found to be.5 C. Therefore, results of two properly onduted tests by the same operator (eah onsisting of the average of three alibration test speimens of the same ement ontent) should not differ

4 Correlation Coeffiient =.974 A-squared = 95 perent aa (/) :I "iii Qi u (/) 17 e Cl : 'iii.!:::! ~ 14 z iii ::: Linear regression through all seventy-two data poi~~~- y = 2.64x ~..... ', '....,,...- Represents 95% onfidene limit within whih average of all data points would fall. Represents 95% onfidene limit within whih any data point would fall... e Cement Content - % FIGURE 2 Phase II alibration test speimens. TABLE 1 SUMMARY OF INTERLABORATORY TEST PROGRAM FOR 5 PERCENT PLUS NO. 4 CALIBRATION TEST SPECIMENS TABLE 2 SUMMARY OF INTERLABORATORY TEST PROGRAM FOR 5 PERCENT PLUS NO. 4 CALIBRATION TEST SPECIMENS Temperature Rise Cement Content Average Maximum Minimum % o o o Standard Deviation within-lab between-lab Temperature Rise Cement Content Average Maximum Minimum Standard Deviation % C C within-lab between-lab Average: Average:

5 TABLE 3 SUMMARY OF INTERLABORATORY TEST PROGRAM FOR 5 PERCENT PLUS NO. 4 TEST SPECIMENS A, B, AND C TABLE 4 SUMMARY OF INTERLABORATORY TEST PROGRAM FOR 5 PERCENT PLUS NO. 4 TEST SPECIMENS D, E, AND F Cement Content Cemenl Conlenl Test Speimen Average Maximum Minimum Standard Deviation Test Speimen Average Maximum Minimum Slandard Deviallon % % % within-lab between-lab % % % within-lab between-lab A D B E F Average: Average: TABLE 5 SUMMARY OF INTERLABORATORY TEST PROGRAM FOR PRECISION DETERMINATION OF CALIBRATION TEST SPECIMENS Phase I Phase II 5% plus No. 4 5% plus No. 4 Standard deviation ol a single test result' Maximum allowable differene e between two test results "' a. I "' ' Maximum allowable differene (?) between the highest and lowest Individual temperature determinations' e:- Standard deviation of a single test result'.65.8 _a -~ Maximum allowable differene 3 between two test results from different "" laboratories 1 A test result is defined as the average of three separate measurements 2 Eah onsisting ol the average of three allbratlon test speimens at the same ement ontent 3 The three individual temperature determinations used in alulating a lest result 4 Numbers in table are In degrees Celsius TABLE 6 SUMMARY OF INTERLABORATORY TEST PROGRAM FOR PRECISION DETERMINATION OF TEST SPECIMENS Phase I - 5% plus No. 4 Phase II - 5% plus No. 4 Test Speimens A, B, and Test Speimen, E, and F Standard deviation of a single lest result' _8 e Maximum allowable differene "' a. between two test results.4.37 <L> ' (?) Maximum allowable differene between the highest and lowest ement ontent determinations e:- Standard deviation of a single test 2 result _o ~ Maximum allowable differene 3 between two test results from "" different laboratories 1 A test result is defined as the average of two separate measurements Eah onsisting ol Iha average ol two ement ontent determinations The two ement ontent determinations used in alulating a test result Numbers in table are In perentage points of ement ontent

6 22 by more than 1.5 C. The range (differene between the highest and lowest) of the three individual temperatures used in alulating a test result should not exeed 3. C. Multilaboratory Preision The multilaboratory standard deviation of a single test result (a test result is defined in this proedure as the average of three separate measurements) has been found to be.8 C. Therefore, results of two properly onduted tests in different laboratories on the same soil-ement mix should not differ by more than 2.3 C. TEST SPECIMENS Single-Operator Preision The single-operator standard deviation of a single test result (a test result is defined in this proedure as the average of two separate measurements) has been found to be.14 of a perentage point of ement ontent. Therefore, results of two properly onduted tests by the same operator (eah onsisting of the average of two ement ontent determinations) should not exeed.4 perentage points of ement ontent. The range (differene between the highest and the lowest) of the two individual ement ontent determinations used in alulating a test result should not exeed.55 perentage point of ement ontent. Multilaboratory Preision The multilaboratory standard deviation of a single test result (a test result is defined in this proedure as the average of two separate measurements) has been found to be.26 of a TRANSPORTATION RESEARCH RECORD 1295 perentage point of ement ontent. Therefore, results of two properly onduted tests in different laboratories on the same soil-ement mix should not differ by more than.74 perentage point of ement ontent. CONCLUSIONS Results of the interlaboratory test program indiate that hanges made to Test Method Q116B by The Bureau of Relamation produe single-operator and multilaboratory preision values that are well within aeptable limits for produing a high-quality end produt. The Bureau of Relamation's revised heat of neutralization test method an be used to aurately and quikly determine the ement ontent of freshly mixed soil-ement ontaining from 3 to 15 perent ement and up to 5 perent plus No. 4 ( mm) sieve size material. The test method does not require separation of the plus No. 4 (4.75-mm) sieve size material for obtaining test speimens, and ement ontent determinations an be made in 15 to 2 min. Required apparatus is durable for field use, the test proedure is easily performed by field personnel, and in field appliations performed to date the method is aurate to within ± 1 perentage point of atual ement ontent. A written test proedure has been prepared for use on Bureau of Relamation projets. The written test method will soon be submitted to ASTM for approval and publiation through its normal onsensus standards proedure. REFERENCE I. Cement Content of Cement Trellted Mllterials (Heat of N1mlralizatio11). Test Method Qll6B- 197, Main Roads Departmenl, Queensland, Brisbane, Australia, Publiation of this paper sponsored by Committee on Soil-Portland Cement Stabilization.