Document npbZg3e8yxLon6oBejZ1VkqJa

2,. UNION CARBIDE INTERNAL CORRESPONDENCE , PROJECT FILE C PY LJ FH C py r--, COMMUNICATION COPY ^ Destroy After Us______ CHEMICALS AND PLASTICS To (N*m.) Division loc*ion Mr. C. K. Wood TECHNICAL CENTER P. O. BOX 8361, SOUTH CHARLESTON, WEI,VJRGINIA 25303 Copy Dt* July 22, 1! originating Oapt. Engineering Copy to Dr. R. R. Ashley Mr. R. L. Aspley Mr. R. C. Buxton * Mr. R. 0. Cavender * Mr. E. G. Dietz * Mr. D. C. WatIH ns Mr. W. C. Hazlebeck Mr. J. H. Howell Mr. R. M. Tfanlrtrtn * Dr. W. R. Manning Mr. J. R. Matthews Mr. W. W. McManus Mr. J. D. Mills * Mr. P. H. Rohr' *Hr. ,(J* A. Snyder Mr. R. H. Steinberg * Mr. C. C. Tanona * . Mr. L. 3. Yandelinder Mr. R. N. Wheeler * (with attachment) subject CB&I Canpany Report on Column Failures Vinyl Chloride Storage Sphere North Charleston Tank Farm J [RECEIVED safes**-)- JUL 231970 1970 L*-JMVHEEUR ^ Dear Mr. Wood: H.v - - -. v.V' V&'gff. A "* ... WsSi - .v -1: .... I- >V\" Attached for your reference is a copy of the long-awaited report from Bridge and ' Iron Company regarding their investigations of the failures'; ^ January 22, 1970, in the legs supporting the vinyl chloride moncmer storage sphere in-North Charleston. - * : ` ''. - ' -' - ' -, 1' ' " According to CB&I, the failures resulted from grout growth caused by the freezing of a very porous grout saturated with moisture. They suggest the growth is accumulative and that unless corrective action is taken befofe next prolonged cold spell, the seams on those columns as yet undamaged will probably burst. , . Jr ' -. '' * * ' ~^J- Whether additional failures will occur in the longitudinal seams is problemat And, if such failures should happen, whether the split seams would actually^pose" a'serious threat to the structural capability of the sphere is at the presj subject to conjecture. x r C- ucc *a& if* -T- > `L. 024364 -ibf` Ix+M'xl jbJ&'a " :* ) & *. Mr. C. K. Wood -2- July 22, 1970 Last February in anticipation of the CB&I report, you and Mr. P. H. Rohr agreed that the Engineering Department must thoroughly analyze these problems and develop recommendations to assure the safe operation of the storage sphere. Furthermore, the Engineering Department routinely specifies concrete-filled legs for fire protection at tanks. The experience -with the sphere in North Charleston has raised questions regarding the wisdcm of this practice; but, we currently suspect that the mortar mix used here was at fault and not the practice of concreting itself. Again, you and Mr. Rohr concurred that the Engineering Department should develop standard specifications for use in fireproofing tank legs with concrete fill. Mr. C. C. Tanona has also endorsed this suggestion,. The target date for completing your investigations and making the recommendations was set for September 1, 1970. This would allow time for corrective measures, if necessary, to be taken on the North Charleston sphere before cold weather. It would also permit the Process Safety and Fire Protection Group to take advantage at an early date of any new standards which may evolve from the studies. Since the principal result of these studies will be standards for general application, you agreed that the cost of the work should be charged to a Technology Engineering account. We have, therefore, closed the Engineering Order and Work Order for the vinyl chloride sphere capital project. My memorandum. Repair of Cylindrical Column Legs, Vinyl Chloride Storage Sphere, North Charleston Tank Farm, dated February 12, 1970, summarizes our experiences with the vinyl chloride sphere, and I have related files and drawings which I can make available to you. However, I recommend that anyone having suggestions and comments should contact you directly regarding additional work which may be required for the sphere legs. Very truly yours, CPW/da C. P. Williams UCC 024365 -C^E0 Chicago Bridge & Iron Company JUL 2 0 197q c''901 WEST 22NO STREET, OAK BROOK, ILLINOIS 60523 July 16, 1970 RECEIVED Union Carbide Corporation Box 8361 South Charleston, west Virginia jul m V/. c. HAZLEBECK Attn: Mr* Hazelbeck Re: Column Failure 48'0 x 75 psig Hortonsphere union Carbide Corporation North Charleston, west Virginia CB&I Contract 69-4255 Gentlemen: Enclosed for your information are one copy each of the strain- gage survey of the support columns and of the petrographic study of the cement grout from within the columns. This is to confirm the information previously presented verbally to members of your organization. In essence, the straingage survey shows that the columns are under internal pressure and, also, that this pressure increases when the temperature drops below freezing. The petrographic study eliminates chemical reactions as the cause of the mortar growth and establishes that the pressure comes from cyclic and/or continuous freezing of a very porous grout saturated with moisture. You will note that Mr. Erlin has included some remarks on the characteristics of the entrapped moisture and how much must be removed to prevent future growth of the grout from freezing. Please note that corrective action must be taken. The growth from the freezing action is accumulative, not elastic, unless the moisture is removed from the grout or other corrective measures are taken, the next prolonged cold spell will probably burst the seams on those columns which have not been damaged thus far. I will be interested in learning what measures are taken to correct this rather unusual situation. Sincerely yours, CHICAGO BRIDGE, & IRON COMPANY CWK/le Attach Manager Foundations Design Oak Brcok Am Coda: 312 654-1700 UCC 024366 CONTRACT 70-5941 STRAIN GAGE SURVEY OF SUPPORT COLUMNS 4S'# x 75 psi Hortonsphere For Storage of Vinyl Chloride Union Carbide Corp. Charleston, West Virginia (Original Contract 69-4255) Author: Project Project T. R. Carbery UCC 024367 TABLE OF CONTENTS LIST OF ILLUSTRATIONS INTRODUCTION SCOPE OF WORK AND PROCEDURE STRESS CALCULATION FORMULA SUMMARY OF RESULTS DAILY LOG LIST OF EQUIPMENT AND MATERIALS STRAIN GAGE READINGS * PAGE III 1 2 8 9 10 12 13 II JOC 02-136$ LIST OF ILLUSTRATIONS FIG. NO. TITLE 1 Orientation of Columns and Location of Columns With Strain Gages 2 Strain Gage Locations On Columns #2 & #3 3 Typical Drilling Pattern For Tre panning Procedure At Active Gage Locations 4 Typical Arrangement and Hook-up of Instrumentation PAGE 4 5 g 7 UCC 024369 INTRODUCTION The purpose of this strain gage survey was to determine the circumferential stresses in two columns of an eight column supported sphere. This sphere was built by Chicago Bridge & Iron Company for Union Carbide Corporation in Charleston/ West Virginia under Contract 69-4255. The sphere is 48' 0 and the product stored is vinyl chloride. The column supports are 28" inside diameter, 0.35" thick and made of A516-Gr.70 steel. The columns are filled with a brick mortar type grout. Three of the columns have split open vertically at various locations along the longitudinal welds. In order to help determine the cause of this failure it was decided to attach strain gages on a column that had failed and one that had not. By using a method similar to * trepanning the circumferential strains in the columns were obtained and from these values the circumferential stresses were calculated. -1- ucc 024370 SCOPE OF WORK & PROCEDURE The two columns selected for strain gaging are shown on Figure #1. Column #3 had split open in the upper and lower sections and the strain gages were attached in the middle section. Column #2 had not failed and the strain gages were attached in the middle section of tuis column too. Column #2 is located between two columns that have failed. On both columns two active gages were attached in similar patterns. One gage was approximately l'-O" above the cir cumferential weld seam between the lower and middle column sections, and the second gage was l'-6" above this weld seam. In addition to the two active gages, a third gage (hereafter referred to as dummy gage) was attached to each column to measure changes in strain resulting from ef fects other than trepanning, such as temperature. This dummy gage was located approximately 3'-01 above the circumfer ential weld seam. All of the gages were oriented with the longitudinal axis in a circumferential direction. See Figure #2 for strain gage location and orientation. The gages were attached using the normal procedure for foiltype gages. A three lead wire system was used to eliminate temperature effects on the lead wire. All gages were covered with synthetic rubber type patch for waterproofing. In addition to the dummy gages mounted on the columns, another dummy gage was attached to a small unstressed piece of steel plate. The gages were connected to a switch box, and the switch box was subsequently connected to a portable battery operated strain indicator. A detailed list of the equipment used is recorded on page 12. A typical wiring hook-up of the instrumentation is shown in Figure #4. After installing the gages the trepanning procedure was begun. The method of trepanning used was to drill a vertical line of holes approximately one inch on each side of the active strain gages. After each hole was drilled the strain on the gage was recorded. The drilling procedure consisted of drilling a pilot hole 1/8" diameter through the steel column plate and then following up with 1/4" diameter hole. The hole pattern was started at the hori zontal centerline of the gage and then alternated above and below this centerline. All steel between the holes was also removed. A typical hole pattern is shown on figure #3. The vertical line of holes was continued until the active gage at the trepanned section showed very little change in the strain reading. Also, after each pair of holes was drilled a strain reading was taken of the dummy gage on the column and the dummy gage on the unstressed steel plate. In addition, to the active and dummy gages being read and recorded, a precision cali brator was connected to the switch box. This was moni tored periodically to be certain th strain indicator was operating properly. 3 ucc 024372 o *,-k Orientation of columns and location of columns with strain gages UCC 4 024373 Vert joint TYPICAL ELEVATION VIEW OP COLUMN #2 Strain gage locations on columns #2 & #3 UCC 024374 5 Approx. 1" LEFT RIGHT Fig. #3 Typical drilling pattern for trepanning procedure at active gage locations ucc 024375 6 Strain indicator `Dial Settings Gage Factor - 2.11 Bridge ------- Half-Quarter Sensitivity----- Midpoint Calibrator .Dial Settings MV/V = +0.5(1000 ue) Switchbox Dial Settings " Gage Resistance --------- 120 Bridge Input------" 2 Arm Fig. #4 Typical arrangement and hook-up of instrumentation. -7- ucc 024376 STRESS CALCULATION FORMULA Circumferential stress (psi) in column, S = E x e where: E = modulus of elasticity for steel = 30 x IQ6 psi e = strain, microinches/inch = initial strain reading minus final strain reading ucc 8 024377 SUMMARY OF :<SSULTG e Initial Strain Final Strain microinches S Gage Location Reading Reading per inch psi Col #2, Upper Col #2, Lower Col #3, Upper Col #3, Lower -426 205 -1150 -1315 -609 10 -1387 -1544 183 5490 195 5850 237 7110 229 6870 Average Stress Col. #2 = -4-Q * 585~ = 5670 psi tension Average Stress Col. #3 = --6 870 gggg pS^ tension It should also he noted that a shift in the strain read ing of the dummy gage on column #3 occurred between the readings on Monday, Feb. 2 and the readings on Wednesday Feb. 4. The indicated strain increase based on the average readings for both days is 21 microinches/inch. Also the atmospheric tem perature which was 40F on Monday was approximately 20F less on the day the final readings were taken. Based upon this indicated strain increase the calculated stress increase is as follows: s = exe - (30 x 106) (21 x 10"6) * 630 psi tension The strain gages on column #2 were not installed until Wednesday and no shift occurred in the dummy gage from the time it was installed until it was removed. There was no sig nificant temperature change during this time. -9- ucc 024378 Daily Log Monday, Feb. 2 - I left Chicago, O'Hare Airport at 7:40 a. M., C. S. T., and arrived in Charleston, West Virginia about 10:45 A. M., E. S. T. After meeting with representatives of Union Carbide Corp., and having lunch I was taken to the job site, arriving at 1:15 P.M.. Pour ironworkers were waiting there to assist in the tre panning operation. A light drizzle was falling inter mittently and the temperature was approximately 40F. About 2:00 P.M., the rain stopped, and I decided to install a strain gage on column #3. The installation checked out satisfactorily and the trepanning was begun by one of the iron workers. The other ironworkers re mained at the site but didn't do anything. At 3:45 P. M., the ironworker stopped the drilling procedure and said it was quitting time. Only five sets of holes had been drilled in the column. Tuesday, Feb. 3 - A heavy snowstorm lasted through most of the day. After discussions with Union Carbide representatives it was decided to put up some temporary shelters around the column legs selected for the strain gage installations. This was done by some Union Carbide personnel. The shelters were built with 2x4 framing and plastic sheeting. No other work was done this day., Wednesday, Feb. 4 - The snow had stopped but the temperature was about 0F. The shelters provided good protection from the wind, and I proceeded with the instal lation of the remaining gages. Fortunately, I had a heat lamp among my equipment and this did provide some extra warmth in the shelters. However, the temperature was still only about 20F and the installations took much longer than they do in warm weather. The metal conditioner and neutralizer 10 UCC 024379 solutions used in one installation procedure had frozen while left in the trunk of my car, and they had to be thawed out. Also a high temperature soldering gun had to be used to connect the lead wires to the gages. A small pencil type soldering iron could not provide enough heat in the cold air. By 11:00 A. M., all the gages were installed and checked out. Two ironworkers then proceeded with the trepanning operation. The first area selected was the one which had been started on Monday, February 2 (the upper gage on column #3). Although I thought the strain reading of this gage probably would have drifted considerably from the final reading made on Monday, February 2., this was not the case. The difference between the final reading on Monday and the initial reading on Wednesday was about a 20 microstrain decrease. Along with this, the change in the reading of the dummy gage on the column had also shifted approximately 20 microstrains. After recording these strain readings the trepanning was continued at this location. When sufficient holes were drilled so that the strain readings did not indicate a significant change, the trepanning was stopped. The pro cedure was then repeated at the lower gage on column #3. After completing all work on column #3 the upper gage location on column #2 was trepanned. The last location to be trepanned was the lower gage on column #2. A final strain reading and resistance to ground of all gages was taken to be certain that all gages were functioning properly and no drifting had occurred. Then all gages and lead wires were removed from the columns. - 11 - UCC 024380 Equipment & Materials Strain Gages Gage Adhesive Gage Waterproofing Lead Wire Strain Indicator Switch Sox Calibrator Megohmmerer Single element foil type with integral terminals, MicroMeasurements EA-06-250 BG-120-W2 Lot #Q-A20AD24 Gage factor - 2.11 - Eastman 910 supplied by MicroMeasurements - Neoprene patch, BLH Barrier "E" - Four conductor cable, #22 AWG (7 x 30), Belden Type 8723 - Portable battery operated Budd P350, serial #1881 - Portable ten channels, BLH #225, CB&I #1129, serial #2392 - Bean #120, Serial #1203 - Portable battery operated, Weston #799 12 UCC 024381 70 -5941 STRAIN GAGE READ/MGS - COLUMN *2 UPPEG GAGS ' _- ' 11 ..f"/. iVjJI 'A 0$rP4/AJi TSestsr+Mct at~m~ (nesmmsj ! r or- M^ *1 2-4-70 (HOME) 426 7/72 968 7000 oo APPRO*. 7 -437 -428 7/70 968 20* z -440 -442 7770 968 3 -445 -446 7772 966 4 -457 -467 7775 963 i 5 -48o -488 7776 966 1 6 -496 500 7775 970 7 -509 573 7778 972 8 -520 -527 7776 970 7000 9 -544 -548 7/78 970 70 -557 -562 //80 968 // -563 -574 7776 966 72 -580 585 7/76 968 /3 592 -595 7/7& 964 74 -600 -602 7780 964 /S -604 -606 7/76 964 /c 60S --6o6 7/74 *966 /7 -607 -- 60Q 7772 968 78 -608 6c9 7770 968 79 -608 -6o9 7774 966 7000 O LO\A SE7Z G. 77o*- No. E/ONT 7.E*r 2-4-70 (none) 7 2 3 4 S 6 7 8 9 70 77 72 73 74 75 76 77 2 05 204 202 200 793 792 787 782 760 777 746 740 72& 7/5 706 99 95 8Z 70 63 56 SO 43 37 37 74 79 76 73 72 72 70 9 70 70 77 72 7/70 7774 1/76 7776 7/72 7776 7/74 7772 7/72 7770 7/70 7772 7/70 7/68 7770 7772 7770 96 a 968 970 972 968 966 968 968 968 970 972 972 970 970 970 970 968 970 7000 ! 7000 7000 OO APPROX- 20* =iii ! 1 1 i1 i \i j ! |1 : iIi1 oo 7 1i ucc 024382 13 70-594/ STRA/N GAGE PEADJNGS - COLUMN *3 UPPER. GA G DATE. //OLE A/a 727GHT LENT 1 2-2'-70 (K/ONE) / 2 3 4 T5 2- 4 -70 s <3 7 8 9 /o /' 72 /3 74 75 76 7? 78 -/> SO -7/54 -7762 -7769 -7772 -7780 - 7797 -7270 -727$ -7222 -7227 -72o4 -7274 - 7226 -7238 -7247 -7258 -7270 -7230 -729/ -7302 -7372 -7320 -7329 -7336 -7344 -7353 -7359 -7J6S -7370 -7374 -7377 -7380 -7382 -7333 -7385 -738& -7387 ON ]'CAL/B4A70A 1 Rfs/seancs atni. i Column SMAiC E 67^0 ! IsretAtN s) ------ (m&OMMS4 c n , 2940 294 Z 968 970 /ooo ! GO APPROXT : 40 2938 ! 970 2938 l 966 2934 968 2954 i 970 7000 CO ' ! Ti 2958 972 7000 GO APPROX 2960 970 2956 ! 970 20 i 2958 2960 968 970 i ! 296Z 2958 2956 2958 97Z 974 7000 970 970 j 1 i ; 2956 972 2958 2960 2962 2964 '968 970 972 970 7000 oo ii l f i1 La\R'E/2 G AGE 7/OiE 77a R/&NT LEPT 2-4- -7o ' (none) 7 2 3 4 S 6 7 8 9 70 77 72 73 74 7S 76 77 78 -A: 1/5 -7320 -7332 -7347 -7354 -7360 -7370 -7372 -7380 -7386 -7393 -7398 -7470 -74/9 -7427 -7433 -7439 -7443 -7458 -7464 -7472 -7483 -7492 -7496 -7503 -7572 -7520 -7526 -7530 -7533 -7536 -754c -7542 -754Z -7544 -7543 -7544 2967 2965 2964 2965 2964 2962 2960 2964 2964 2962 2960 2958 2958 2956 2958 2956 2956 2958 2958 972 974 975 9 74 972 972 970 972 970 974 972 974 972 972 972 972 970 970 972 /OOO /OOO 7000 7000 oo o * APPROX. 20 i i 1j t i - r1 ucc 02-4383 SI1 SKOKIE BOULEVARD NORTHBROOK. ILLINOIS 00002 BERNARD ERL.IN MATERIALS ANO CONCRETE CONSULTANT PETROGRAPHER . GEOLOGIST - MINERALOGIST OFFICE. (312) 272.7730 IW.i (312) 037.0201 A PETROGRAPHIC STUDY OF MORTAR FROM FAILED COLUMNS m CHICAGO BRIDGE AND IRON COMPANY ******** SUMMARY AND DISCUSSION. Microscopical, infrared, and chemical studies showed that the grouts had been made using mixtures of portland cement, natural sand, mineral and chemical admixtures containing fly.ash, lignosulfonate, and triethanolamine, and a relatively high mixing water content. The water-cement ratio had been higher for the grout from the failed column and was denoted by extensive bleeding channels. An apparent retardation to hydration caused by either or combinations of retarding admixtures and low temperatures during placement of the grout was manifested by the mineralogy of the cement hydration phases. Additionally, secondary compounds hsui formed in voids and on fracture surfaces. There was no evidence in the specimens that adverse chemical reactions including alkalisilica reactivity or sulfate attack had been operative. The damage to the columns is interpreted to . have resulted from cyclic freezing of grout saturated with moisture. The source of the moisture was the mixing water which greatly exceeded the demand of the portland cement for water -- the excess was entrapped within the sealed columns. ucc 024384 T BERNARD ERUN - MATERIALS AND CONCRETE CONSULTANT Based, upon laboratory studies of the specimen from the failed column, and assuming that the specimen was representative of that column, grout in the column contained about 10,8 percent evaporable water at 95C# . To minimize, if not eliminate completely, future damage by cyclic freezing, it will be necessary to reduce the moisture content of the grout below a critically saturated level. The level necessary to reduce the triaxial dilation incidental to freezing, varies de pending upon the character of the cement paste and aggregate and hence, is different for different concrete or mortar. Below a paste relative humidity of about 80 percent, there is no freezable water and absence of damage by freezing is virtually assured. That humidity, however, requires relatively long periods to attain. Certainly, a higher moisture content may also be acceptable, however, the exact moisture level cannot be calculated. For green concrete, an upper saturation level of 97 percent has been shown as the value limiting damage by freezing. Certainly, a relative humidity below that level should be accept ble for hardened concrete. Thus, if the grout dries to an internal relative humidity of 97 percent or less, damage by cyclic freezing should be minimal. Data available from studies of 6 inch diameter mortar cylinders having 0.6 water-cement ratios show that about 125 days were required to bring the relative humidity at the centers of the specimens to between and 100 percent. Storage for the cylinders was at 50 percent R.,H. and 75F. The drying period for the encased grout will be longer because of the great reduction in the surface area from which drying can occur. Forced drying using heat will materially reduce the drying time. Temperatures gradually increased to about 180F should be used. Because a drying period * necessary to attain a desired humidity cannot be predetermined, it is advisable to monitor the moisture level within at least one column. Several monitoring UCC 024385 BERNARD ZRLiN - MATERIALS and concrete consultant i systems that employ sealed wells and moisture sensitive probes are available. ****-** imoamfia Reported herein are the results of analytical studies for specimens from two grouts used for filling sealed steel tubular columns. The studies were requested by C. W. Kemp of Chicago Bridge & Iron Company, to resolve the cause for splitting of several columns. Mr. Kemp reported that the columns had been filled last October with a grout mixture composed of portland cement and fine river sand -- no admixtures were reported. Eight 48 inch diameter columns were involved -- 3 of the columns had split apparently during January. The specimens forwarded for the studies were from a failed column (Specimen No. 2) and a column displaying no external damage (Specimen No. 1). Failure occurred during sub freezing temperatures. STEGIMS MB-Sm>IB Heceived for the studies were two 2 inch diameter cores identified as 1 and 2. The former came from a failed column; the latter from a sound column. Core 2 was sealed in a plastic bag and was saturated -- Core 1 was not housed in a sealed container and was dry. Each specimen was examined using microscopical methods and portions were analyzed using wet chemical and in frared spectroscopical methods. The studies were directed toward identifying original constituents of the grout and subsequent affects, both physical and chemical, that might have contributed to the failure. Additionally, sections from each specimen were tested in drying and cyclic freezing environments to explore the effects of those exposure on the volume stability. -3- 1i UCC 024386 BERNARD ERLIN - MATERIALS and CONCRETE CONSULTANT THE STUDIES Microscopical Examinations - Both specimens were similar in compositional properties and had been made using Portland cement, fly ash, and a natural sand composed primarily of quartz and feldspar, and lesser amounts* of shale, mica, and soft coal. Both mortars had been made using a high water-cement ratio - Specimen 2 by comparison to Specimen 1 had a much higher water-cement ratio. The higher watercement ratio, in part, had resulted in extensive channeling of bleed water and voids formed by the channeling re mained. Those areas and entrapped voids contained large thin hexagonal platelets of calcium hydroxide, and deposits of ettringite, Similar deposits less extensively developed occurred in Specimen 1. Calcium hydroxide also occurred as relatively large crystals intergrown with the paste of both cores. + Paste in both cores was relatively porous, soft, and composed of relic portland cement, glassy spheres charact eristic for fly ash, calcium hydroxide and other hydration products of the portland cement. Ettringite was present, however, it was not intergrown with the paste and Is considered a secondary product not reflecting sulfate attack. There was no evidence that adverse chemical reactions . had occurred, or that aggregate particles had dilated. Some coal particles displayed internal cracking characteristic for shrinkage. Infrared Studies - Portions from each specimen were processed and analyzed for organic materials used in chemical admixtures for concrete. Identified in extracts of the specimens were a) lignosulfonate in amounts inter preted to be about 5 times typical additions for Specimen 1, and 3 times typical additions for Specimen 2, and b) triethanolamine at relatively high concentration^. Chemical Studies - Portions from each specimen were analyzed for sulfate using a gravimetric method. Specimen 1 was determined to contain 0.51 percent sulfate (as SO-) by weight of mortar and Specimen 2 contained 0.49 percent sulfate. Those amounts are within the sulfate levels one would anticipate to be contributed by the portland cement. Thus, that data would confirm the microscopical Interpretations that the concrete had not undergone sulfate attack. UCC -4- 024387 QERNARO ERUN - MATERIALS and CONCRETE CONSULTANT Volume Stability.. Studies - Two sections from each specimen were processed and one section from each core was stored at 100F in sealed plastic bags* The other sections viere cycled from ambient room temperature to OC in sealed plastic bags. The processing included saturating the sections by submerging them in water and placing them in a saturated surface dry condition in plastic bags which were then sealed. Initial and periodic length measurements were made for the specimens stored in each environment. Both specimens stored at 100F expanded for several days as would be anticipated due to thermal response to the higher temperature. Subsequently, they began to shrink and after 17 days of storage shrank 0.11 percent. That decrease is attributed to drying of the specimens which probably occurred because of leaks to the plastic bags. Progressive expansions have occurred *for the specimens cycling from ambient to freezing temperatures. After 17 cycles, expansions were 0.025 percent for Specimen 1 and 0.035 percent for Specimen 2. length change data for the specimens are sljown in the attached Table. Kolsture Studies - Portions from Specimen 2 were heated at 95C to constant weight to determine the degree of saturation. Assuming that the grout was saturated similarly in the columns (it was reported that the grout was moist) and that coring had not affected the specimens, then evaporable water at that temperatures was equivalent to 10.8 percent by weight of grout. ucc -5- 024388 BERNARD SRLIN MATERIALS and concrete consultant TABLE - Length change data for specimens stored at 100F and specimens cycling at ambient and freezing temp eratures . H CM Specimen Storage Condition 100F 100 F Length Change as Percent of Initial Length, for Periods Shown 7. da. 17 da. +0.025 +0.015 -0.040 -0.015 -0.11 -0.12 Cyclic freezing Cyclic freezing +0.015 +0.015 +0.020 +0.020 +0.025 +0.035 UCC 024389 -6-