US Army Corpsof Engineers®Cold Regions Research &Engineering Laboratory

Expedient Low-TemperatureConcrete Admixtures for the ArmyCharles J. Korhonen November 1999

20000328 040MTC QUALMT INSPE=.D, 3

04/120/00 14:30 FAX 603 646 4712 CRREL Library U•001

COLD REGIONS RESEARCH & ENGINEERING LABORATORYLisa HoffmeisterTechnical Library tel: 603-646-433872 Lyme Road fax: 603-646-4712Hanover, NH 03755-1290 email: erhoff@crrel41.crrel.usace.army.mil

4/20/00

TO: Judah Dnnis

FAX: 703-767-9089

RE: report number correction

Number of pages: 1

Message

Judah,

I hope I spelled your name correctly. If not, let me know (and I apologize).

Please change the following report numbers from:

CRRL-SP-99-1 7 (AD-A375 241) - 6)6'6VZTo

CRREL-SR-99-17

And

CRREL-SP-99-1 8 (AD-A375 238) - 0 )'2~ ' ~.

To

CRREL-SR-99-1 8

Thank you.

REPORT DOCUMENTATION PAGE Form Approved0MB No. 0704-0188

Public reporting burden for this collection of information is estimated to average I hour per response, including the time for reviewing instructions, seerching existing date sources, gathering and mnaintaining thedate needed, and completing and reviewing this collection of information. Send comments regarding this burden entimate or any other aspect of this collection of information, including suggestions for reducingthis burden to Department of Defense, Washington Headquarters Services, Directorate for tnformation Dperations and Reports (0704.0188), 1215 Jefferson Davis Highway. Suite 1204, Arlington, VA22202-4302.Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently validDMB control number. PLEASE DD NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.1. REPORT DATE (oD-MM-YY) 2. REPORT TYPE 3. DATES COVERED (From - To)

November 19994. TITLE AND SUBTITLE 5a. CONTRACT NUMBER

Expedient Low-Temperature Concrete Admixtures for the Army 5.GATNME

5c. PROGRAM ELEMENT NUMBER

4A762784AT426. AUTHOR(S) 5d. PROJECT NUMBER

5e. TASK NUMBER

Charles J. Korhonen5f. WORK UNIT NUMBER

AT42-TO-O1 l7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT

NUMBER

U.S. Army Engineer Research and Development CenterCold Regions Research and Engineering Laboratory72 Lyme Road Special Report 99-17Hanover, New Hampshire 03755-1290

9. SPONSORINGIMONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR / MONITOR'S ACRONYM(S)

Office of the Chief of Engineers 11. SPONSOR I MONITOR'S REPORTWashington, D.C. 20314 NUMBER(S)

12. DISTRIBUTION/I AVAILABILITY STATEMENT

Approved for public release; distribution is unlimited.

Available from NTIS, Springfield, Virginia 22161.

13. SUPPLEMENTARY NOTES

14. ABSTRACT

Over 50 chemical combinations were studied for their effect on the strength gain of mortar when it is below freezing. Thechemicals studied were limited to those that are readily available most anywhere, but that are not marketed as admixturesfor concrete. Four admixtures were developed from these chemicals that allowed mortar to gain full strength at -10'C asrapidly as normal mortar cured at temperatures between 5 and 1 01C. A host of other chemicals allowed mortar to recoverfull strength when cooled to the point that cement hydration ceased or that allowed mortar to continually gain modeststrength while at low temperature with only minimal loss of ultimate strength. Guidance is provided for expedient fielduse-for concrete not expected to last more than 5 years.

15. SUBJECT TERMSAntifreeze mixture Expedient constructionCold-weather concrete Low-temperature construction

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF RESPONSIBLE PERSONOF ABSTRACT OFIPAGES

a. REPORT b. ABSTRACT c. THIS PAGE 19b. TELEPHONE NUMBER (include area code)

Unclassified Unclassified Unclassified UL 20

Standard Form 298 (Rev. 8-98)Prescribed by ANSI Std. 239.18

Abstract: Over 50 chemical combinations were stud- and 1 0°C. A host of other chemicals allowed mortar toied for their effect on the strength gain of mortar when it recover full strength when cooled to the point thatis below freezing. The chemicals studied were limited cement hydration ceased or that allowed mortar toto those that are readily available most anywhere, but continually gain modest strength while at low tempera-that are not marketed as admixtures for concrete. Four ture with only minimal loss of ultimate strength. Guid-admixtures were developed from these chemicals that ance is provided for expedient field use-for concreteallowed mortar to gain full strength at -10OC as rapidly not expected to last more than 5 years.as normal mortar cured at temperatures between 5

How to get copies of CRREL technical publications:

Department of Defense personnel and contractors may order reports through the Defense Technical Information Center:DTIC-BR SUITE 09448725 JOHN J KINGMAN RDFT BELVOIR VA 22060-6218Telephone 1 800 225 3842E-mail help@dtic.mil

msorders@dtic.milWWW http://www.dtic.mil/

All others may order reports through the National Technical Information Service:NTIS5285 PORT ROYAL RDSPRINGFIELD VA 22161Telephone 1 703 487 4650

1 703 487 4639 (TDD for the hearing-impaired)E-mail orders@ntis.fedworld.govWWW http://www.ntis.gov/index.html

A complete list of all CRREL technical publications is available fromUSACRREL (CEERD-IM-HL)72 LYME RDHANOVER NH 03755-1290Telephone 1 603 646 4338E-mail techpubs@crrel.usace.army.mil

For information on all aspects of the Cold Regions Research and Engineering Laboratory, visit our World Wide Website:

http://www.crrel.usace.army.mil

Special Report 99-11 a.US Army COrpsof Engineers®Cold Regions Research &Engineering Laboratory

Expedient Low-TemperatureConcrete Admixtures for the ArmyCharles J. Korhonen November 1999

Prepared forOFFICE OF THE CHIEF OF ENGINEERS

Approved for public release; distribution is unlimited.

PREFACE

This report was prepared by Charles J. Korhonen, Research Civil Engineer, Civil Engineer-ing Research Division, Cold Regions Research and Engineering Laboratory, U.S. Army Engi-neer Research and Development Center.

The investigation was funded by DA Project 4A762784AT42, Military Engineering Technol-ogy, Cold Regions Theater of Operations; Work Package 162; Work Unit AT42-TO-011, ExpedientCold-Weather Concrete Admixtures. Technical review of this report was provided by DeborahLoiselle, Research Supervisor, New Hampshire Department of Transportation, and Dr. AsokSarkar, Research Engineer, The University of Dayton, Dayton, Ohio. The author acknowl-edges the diligent work of Sherri Orchino, Civil Engineering Technician, CRREL, for herlaboratory expertise in conducting these tests.

The contents of this report are not to be used for advertising or promotional purposes.Citation of brand names does not constitute an official endorsement or approval of the use ofsuch commercial products.

CONTENTS

Preface ....................................................................................................Introduction ............................................................................................ 1.Effect of temperature on strength .................................................................. 1.Test program............................................................................................. 2

Materials .............................................................................................. 2Mixing ................................................................................................. 2Sample preparation and curing................................................................... 2Test method........................................................................................... 3

Results and discussion................................................................................. 3Accelerators.......................................................................................... 3Freezing point depressants ........................................................................ 4Combinations........................................................................................ 5Other ways to view antifreeze admixtures ..................................................... 6

Conclusions and recommendations ................................................................. 6What we still don't know ............................................................................. 8Literature cited.......................................................................................... 9Appendix A: Performance data.......................................................................1Abstract .................................................................................................. 17

ILLUSTRATIONS

Figure1. Relationship between chemical dosage and freezing point.............................. 42. Best strengths at -100 C.......................................................................... 5

TABLES

Table1. Effect of temperature on strength gain of control mortar................................. 22. Early age strength from the best accelerators............................................... 33. Mortars that recovered more than 95% strength after being cured at -20'C for

28 days and then at 200C for an additional 28 days.................................... 64. Admixtures that will produce strength at -100C at least as fast as normal concrete

cured at P0C................................................................................... 75. Everyday chemicals that allow concrete to gain at least 75% strength after being

at low temperature for 28 days followed by room temperature for 28 days.....7

iii

Expedient Low-Temperature Concrete Admixtures

for the Army

CHARLES J. KORHONEN

INTRODCTIONon early age strength gain. It did not considerCurrently, portland cement concrete cannot be their effect on paste shrinkage, aggregate degra-

placed at below-freezing temperatures without dation, reinforcing steel corrosion, or concretesome form of thermal protection. Therefore, Field durability. (These studies should be conducted ifManual, FM 5-428, Concrete and Masonry (U.S. the concrete is intended for long-term use.) TheArmy 1998), directs that freshly placed concrete objective of this study was to develop guidance tomust be protected from freezing by keeping it enable military engineers to design expedientabove V0C with insulation or heated enclosures cold weather concrete admixtures, from chemi-for a minimum of 3 days or until it has sufficiently cals that are readily available but not marketedcured to gain strength and serve its intended pur- expressly as concrete admixtures, for field expe-pose. On the battlefield or in an emergency, when dient use. Expedient concrete is defined as con-concrete must be placed regardless of the temper- crete that is intended to have service life up to 5ature, engineers may not have access to insula- years.tion, let alone heated enclosures. If concrete mustbe placed in the cold, it may freeze and the results EFFECT OF TEMPERATURE ON STRENGTHcan be disastrous.

An alternate approach to winter concreting, Portland cement concrete, in its simplest form, isand one not yet recognized by standard practice, composed of cement, aggregate, and water. Ais to use antifreeze admixtures. These are chemi- chemical reaction between the cement and watercals that depress the freezing point of water and forms a gel that binds the aggregate together andaccelerate the hydration rate of cement when the gives concrete its strength. As long as the cementinternal temperature of the concrete is below 00C. is wet, the concrete continues to gain strength at aCommercial standards do not permit the use of rate that is strongly influenced by temperature.antifreeze admixtures because their long-term Common belief is that hot weather is the besteffect on concrete is unknown. However, two time to cast concrete because concrete sets uprecently completed studies (Korhonen and Brook faster when it is hot than when it is cold. Table 11996, Korhonen et al. 1997b) concluded that cer- supports this belief by showing that mortar gainstain chemicals could protect concrete from freez- 77% more strength in 1 day when cured at 500Cing without causing detrimental side effects. But, than it does when cured at 20'C. Getting thispartly out of resistance to change and partly for much extra strength in 1 day can save an engineerother reasons, antifreeze admixtures have not several days of extra expense in labor and in wait-been commercialized and are -not in general use. ing time to reuse the forms. However, the

This report details a laboratory investigation strength benefits are only temporary. As can bein search of chemicals to allow portland cement seen, the early age strength increase caused byconcrete to resist freezing during initial curing, high temperatures virtually disappears by day 7.This search targeted chemicals, other than those Moreover, the mortar ultimately becomes 28%specifically marketed for concrete, for their effect weaker when cured at 50'C than when cured at

Table 1. Effect of temperature on strength gain* of freezing point depression of mix water and thecontrol (admixture-free) mortar. strength gain of mortar as a function of time and

Temp Strength?' temperature.(-C) 1 day 3 days 7 days 14 days 28 days 56 days

50 17.7 33.6 58.1 69.3 75 77 Materials40 16 38.1 69.7 79.2 85 85 This study used mortar as a rapid way to eval-30 13.1 36.6 72.1 87.3 94 93 uate various chemicals. Mortar, instead of con-20 10 30 70 90 100 105 crete, simplified mixing operations, reduced10 3.7 15 49 75.6 97 110 material handling, and permitted smaller test5 2 14 30.1 56.7 83 100

-5 1 6.7 14 16.6 20.3 46.8 specimens, which meant less material had to be-10 0 0.2 1.7 2.7 4.2 51.1 taken to the landfill. In addition, mortar closely-20 0 0 0 0.2 0.8 37.3 simulates concrete because it has the same paste*The strengths for 20, -5, -10, and -20'C are results pro- to aggregate transition zones, except that theduced at CRREL on 0.39 w/c mortar. The others are expected aggregate is smaller. The mortar was non-air-strengths according to ACI (1988). entrained,* used Type I portland cement at a

tStrengths are given as percentages of the 28-day strength of 1:2.75 ratio with sand, and was made with water-control mortar cured at 20'C. The mortars were cured at thetemperature shown in the left most column for 28 days, then to-cement (w/c) ratios between 0.39 and 0.45 (thethey were cured at 20'C for the remaining 28 days. majority of the testing was at 0.41 w/c). The sand,

sieved for use in concrete, had a bulk specific

20'C. Thus, hot weather might allow the project gravity (saturated surface-dry) of 2.65 and a

to go faster but it can result in permanent strength moisture absorption of 1.1%. The mixing water

loss. was from the taps at CRREL.

In the long run, concrete becomes stronger Mixingwhen cured at temperatures that are low but Mixing of the mortar followed standard labo-above freezing. Their drawback is that they slow ratory procedures. The mortars were mixed in adown strength gain, which can adversely affect Hobart mixer according to ASTM (1987) Standardconstruction schedules. If concrete can be mixed C305 with minor modifications. The chemical ad-and protected so that its temperature can be kept ditives were dissolved in the mixing water beforeat V0C, preferably at 10'C, strength gain can be the entire solution was placed into a damp mix-rapid enough to avoid too much delay. Table 1 igbowl. The mixer wstreonalwspee

shos tatmorar urd a 10C eveops63and for 30 seconds while the cement was added. The50% less strength compared to mortar cured at mixer was stopped, the sides of the bowl were20'C for 1 and 3 days, respectively. By 56 days, scraped down, and mixing was resumed for 45this strength deficit turns into a strength surplus seconds while the sand was added. The mixer

of narl 5%.Comaredto 0'Cit ecoms a was briefly stopped, changed to medium speed,33% strength surplus. Curing at 5OC produces and run for 30 seconds. Then, the mixer wassimilar results. Thus, cold weather can be the best stpefo1/2mnesborfialbigrutime to cast concrete, col stpediu for e 11o2 minutes tefoefnly.en u

However, too much codleads to problems. If atmdu spefo1mit.concrete freezes at an early age, it can be per- Sample preparation and curingmanently damaged. As Table 1 shows, the 1 -day The mortar was mixed at room temperature.strength of -5'C mortar is only 10% that of mortar Immediately after mixing it was cast into 50.8- xcured at 201C; no appreciable strength developed 101.6-mm plastic cylindrical molds, vibrated on aat -10 or -201C. Though hydration occurs at tem- table to ensure consolidation, capped with plasticperatures as low as -20'C, albeit slowly, less than lids, and stored at 20, -5, -10, and -20 0 C (somehalf of the concrete's strength can be recovered by mixes were not stored at -20'C to simplify test-thawing. Thus, cold weather protection is ing). The samples were placed into their respec-required to achieve strong concrete when the tive curing rooms within 30 minutes after watertemperature dips below 00C.

TEST PROGRAM *Non-air-entrained mortar was used intentionally to deter-mine the effect additives have on strength development with-

The test program defined the effects of various out having to contend with the variability of entrained air.chemicals, singly and in combination, on the Entrained air is recommended for field applications.

2

first came in contact with the cement. They together. Rather than discuss each table in detail,remained sealed until being tested for strength. I summarized them into the tables and graphs

that follow and referenced them, as necessary, inTest method the text below. The reader is encouraged to re-

Three samples of each mix from each tempera- view Appendix A for additional detail.ture were tested in uniaxial compression accord- In reading this report, it is important to noteing to ASTM (1997) Standard C39 at 7, 14, 28, and that all strengths are referenced to the 28-day56 days (some samples were not tested at 14 or 56 strength of control mortar cured at 20'C. Table 1days owing to scheduling problems). Each samn- was organized in this manner. Likewise, Appen-ple was capped with unbonded neoprene held dix A lists all strengths as percentages.within a steel-retaining cup according to ASTM(1993) Standard C 123 1. The samples from the cold- Acceleratorsrooms were allowed to thaw to V0C before being One function of an antifreeze admixture is totested. accelerate the hydration rate of portland cement.

Any increase in the early strength development

RESULTS AND DISCUSSION shortens the curing and protection periods neces-sary for the concrete to attain the desired

Appendix A shows freezing point and strength strengths. Table Al highlights chemicals thatdata for over 50 chemicals. Although many more force mortar to gain strength faster than it nor-could have been studied, I focused on a small mally would. Table 2, extracted from Table Al,portion of chemicals known to be compatible shows the best results.with concrete. They are grouped into Tables Al It is clear that the seven chemicals shown inthrough A4 by their ability to accelerate the Table 2 increased the early strengths of mortar.strength gain of mortar, by their ability to depress They produced 7-day strengths at 20'C that manythe freezing point of mixing water, by those that times exceeded that of control mortar. In fact,are liquid, and by those that were combined compared to Table 1, they all far outperformed

Table 2. Early age strength from the best accelerators from Appen-dix A.

Percent by Freezingweight of point 7 days*

Chemical cement ('C) 20-C -50C -10-C -200C

Calcium chloride 1 -2.5 108.5 10.5 1.8 na2 -3.8 109.9 19.0 3.6 na3 -6.2 108.5 30.0 5.3 na

Calcium bromide 4 -2.9 108.4 4.9 3.2 0.86 -4 109.4 11.0 6.2 1.88 nat 107.5 4.6 1.8 0.3

Fertilizer (calcium 3 -5.3 114.3 3.8 3.4 3.4nitrate) 6 -6.7 107.9 3.2 3.1 1.9

9 -8.3 96.7 6.8 6.4 5.3

Calcium nitrite 2 -3.5 87.1 13.3 2.7 0.76 -5.7 106.0 44.8 4.4 1.19 na 124.4 67.8 53.2 0.5

Calcium chloride 3 -1.5 92.6 16.9 8.2 5.0(deicer) 6 -2.25 106.3 22.7 9.4 4.2

9 -4 93.9 43.1 34.9 16.4

Calcium acetate 2 -3 94.8 12.1 1.0 na4 -4.7 94.4 2.7 2.0 na6 -6.5 102.4 6.7 3.1 na

Calcium formate 2 -3.3 96.2 14.4 2.0 0.24 -4.5 91.1 17.9 3.3 0.46 -5.5 91.7 7.3 3.3 0.7

*Strengths are given as percents of 28-day control mortar cured at 20'C.fTests were not conducted, or information was not available.

3

the control mortar at 50'C. Except for calcium nitrite were the most active chemicals studied.chloride and calcium bromide, the amount of They produced 7-day strengths at -10'C thatstrength increase depended on the amount of were comparable to control mortar cured betweenchemical added to the mortar. The calcium chlo- 5 and 10'C. Since accelerators are only one com-ride and calcium bromide produced approxi- ponent of an antifreeze admixture, these resultsmately the same 7-day strengths despite differ- are remarkable. Interestingly, too, the lowestences in doses. measured freezing points of the mix waters were

The best known cement hydration accelerator -6.2 and -5.7'C for the calcium chloride and calci-is calcium chloride. As Table 2 shows, it provides um nitrite, respectively. These are not insignifi-more strength gain at a given dosage than the cant from practical considerations as the controlother accelerators. It is also the cheapest of all the mortar froze at -1.3'C.chemicals tested in this study. However, a majordrawback of calcium chloride is its tendency to Freezing point depressantscorrode metal. As a result, calcium chloride is not The second and final function of an antifreezerecommended by current standards beyond 2% admixture is to depress the freezing point ofby weight of cement in reinforced concrete and is water. This function depends on a colligativenot recommended at all in pre-stressed concrete. property called molality-the number of parti-But, for expedient purposes where long-term cles (moles) that dissolve into 1 kg of water. Thiseffects are less important or for non-reinforced is a group of chemicals that are generally weakstructures where corrosion is not an issue, higher accelerators or retarders, and includes both elec-dosages are acceptable. Up to 4% was included in trolytes and nonelectrolytes. Electrolytes are ma-this study For practical purposes, the amount of terials that dissociate into more than 1 mole ofaccelerator that can be added to concrete is limited ions per formula weight, whereas nonelectrolytesby how early the concrete stiffens before interfer- remain in molecular form. Examples of electro-ing with placing, consolidating, and finishing lytes are sodium chloride and sodium acetate andoperations. Since ready-mix concrete is often pro- of nonelectrolytes are alcohols and glycols. Theyduced at temperatures around 20'C during the work on the principle that solute particles lowerwinter, the accelerating potential of chemicals the vapor pressure and, thus, the freezing point ofwas judged at this temperature. Based on this water.study, the optimum dosage of accelerators Figure 1, developed from Appendix A data,appears to be between 4 and 8% for concrete shows that, generally, the greater the number ofcured at 20'C. solute particles, the lower the freezing point. This

At lower temperatures, accelerator dosages relation is independent of the type of chemicalcan be higher. Calcium chloride and calcium used, provided the chemical remains soluble at

0 A I I I I III I I I I

-5AA

10 A-10

ES-15

N

-20

0 2 4 6 8 10 12

Concentration (molality)

Figure 1. Relationship between chemical dosage and freezing point.

4

low temperatures. As seen, there is a scattering of strengths developed by antifreeze mortars curedthe data points about the least-squares regression at low temperature to those developed by normalline drawn through them. This scattering is prob- mortar cured according to current guidelines. Inably caused by variations in attractive forces practice, fresh mortar must be kept at or abovebetween particles and in solubility from chemical 5'C to ensure against freezing and assure a rea-to chemical. The least-squares line reveals that sonably rapid strength gain. Therefore, theeach mole (mol) of solute reduces the freezing strength-gain curve for control mortar cured atpoint (FP) of the water by 1 .76'C. Further, it V0 C is shown in Figure 2 as the benchmark for theshows that admixture-free mortar should freeze antifreeze mortars in this study. The curve forat -1.28'C. (The measured freezing point of the 20'C mortar is shown for comparison. Anycontrol mortar was -1.30C.) admixture that caused mortar, cured below 00 C,

to gain strength at least as rapidly as control mor-FP = -1.76 mol - 1.28. (1) tar cured at 5'C was deemed acceptable.

Figure 2 shows four mortars that produced atInterestingly, the relationship between molali- least benchmark strengths at -10'C. Quite a few

ty and freezing point remains quite good, even at of the chemicals produced excellent results atthe highest dosage. Thus, eq 1 is a good first esti- -5'C, while none passed the test at -20'C. Themate of expected freezing point whenever the best single chemical was sodium nitrite. At a dosechemical composition of the admixture is known, of 9% by weight of cement, this chemical pro-However, it is always best to experimentally duced strengths at -10'C that not only exceededdetermine the freezing point because, as noted, the benchmark, they nearly equaled the strengthsthe colligative behavior of chemicals can vary of control mortar cured at 20'C between 14 and 28from that predicted by eq 1. days. Thereafter, the sodium nitrite outper-

formed the 20'C control mortar. This immediatelyCombinations suggested that sodium nitrite, which is classified

The basic antifreeze admixture utilizes both as a freezing point depressant in Table A2, couldaccelerators and freezing point depressants. be improved further by combining it with anOther chemicals, such as plasticizers, retarders, accelerator-something from Table Al.and air entrainers, can be used, but this study did Two accelerators provided that improvement.not evaluate them. My objective was to find Potassium carbonate and calcium nitrite eachchemical combinations that promote "adequate" affected sodium nitrite's performance at -10'Cstrength gain while the internal temperature of differently. (Calcium chloride was tested in thethe mortar was below O0C. form of a commercial deicer along with sodium

One way to judge adequacy is to compare the nitrate as a fertilizer. They did not perform as

160 I I 1 1 1 1 1

140

120

100

75, 8

60 -- ControlI20'CI - Control 5"C

40 S

20 NS

0 5 10 15 20 25 30 35 40 45 50 55 60

Time (days)

Figure 2. Best strengths at -100C.

5

well as expected. See Table A4 for their perfor- Other ways to view antifreeze admixturesmance.) Calcium nitrite initially caused the early If early age strength is not critical, another wayage strengths to be somewhat delayed, but they to judge the usefulness of antifreeze admixtureswere still well above the benchmark level. The is by whether they ultimately allow mortar tostrengths beyond 14 days became significantly gain full strength even though freezing takesgreater than those of the 20'C mortar, and greater place during curing. The practicality of this is thatthan those of the single component sodium nitrite a concrete structure could be cast during coldfor that matter. This combination appears to be a weather and then later returned to when thelong-term strength enhancer. The potassium car- weather turns warm. In this light, we see in Tablebonate, on the other hand, caused the mortar, at 3 that, even though none of the chemical dosages-100C, to act as if it was at 200C at all ages. This promoted strengths acceptable to the previousadmixture effectively canceled the effect of cold criteria while being cured at -20 0C, 12 of themweather. However, sodium is an alkali that can allowed essentially full recovery of strengthcause certain siliceous aggregates to swell when thawed. Thus, antifreeze admixtures candestructively within concrete. More work should protect concrete against frost damage down tobe done to find alternative freezing point chemi- -200C, perhaps lower, even though strength devel-cals that are benign to aggregate in the long-term ops very slowly at that temperature.or to find ways to mitigate the negative effect of Another viewpoint is to consider that we canadded alkalis. For the short-term, these chemicals expect normal concreting practices during theare fine as expedient admixtures. summer to result in more than a 25% loss of ulti-

Sodium sulfate, when combined with sodium mate strength (Table 1). If one were to hold winternitrite, forms the basis of the antifreeze admixture concrete to that same expectation, then many ofpatented by the U.S. Army in 1993. As Figure 2 the chemicals tested in this study would qualifyshows, it produces excellent strengths up to 14 as acceptable antifreeze admixtures.days when cured at -10'C but seems to providelittle benefit thereafter (it exceeds benchmark CONCLUSIONS ANDstrengths at all ages when cured at -5'C). For ex- EO MN AINpedient purposes, this combination of chemicals REO MNAI Sis still viable, as most severely cold weather does Until now, insulation, supplemental heating,not last more than a few days in most parts of the and tenting have been the only ways to protectworld. The chemicals are readily available most fresh concrete from freezing during cold weather.anywhere, and destructive alkali reactions take Though effective, they are costly in labor, mater-10 or more years before they become a problem. ials, and money-and sometimes they are not

Table 3. Mortars that recovered more than 95% strength afterbeing cured at -20'C for 28 days and then at 20'C for an addi-tional 28 days.

Percent strengthPercent Freezing of control mortar

by weight point cured for 28 daysChemical of cement (10C at 201C

Calcium bromide 4 -2.9 105.76 -4.0 127.8

Calcium nitrite 6 -5.7 118.8Calcium chloride (deicer) 6 -2.25 110.3

9 -4.0 110.5Calcium magnesium acetate 8 -8.2 96.4Magnesium chloride 6 -11.5 96.9Car antifreeze-ethylene glycol 6 -6.6 101.0Calcium chloride deicer/sodium 3/3 -5.1 98.5

nitrate fertilizer 4.5/1.5 na* 99.5Calcium chloride deicer/potash

fertilizer 4.5/1.5 -6.6 98.2Fertilizer (calcium nitrate)/car

antifreeze (ethylene glycol) 1.5/4.5 -5.2 96.5

*Tests were not conducted, or information was not available.

6

Table 4. Admixtures that will produce strength at low temperature, but it recovers full strength-100 C at least as fast as normal concrete cured at 5°C. once warm weather returns. Thus, antifreeze

Dosage Water to admixtures are sort of an insurance policy againstby wgt of cement cement ratio unexpected freezing.

Admixture Chemicals (%) (w/c) The above discussion assumes that concrete

1 Sodium nitrite 6.0 0.45 cast during cold weather must attain significantPotassium carbonate 0.06 early or full late-age strength, or both, to be ser-

2 Sodium nitrite 6.0 0.48 viceable. However, attaining full strength maySodium sulfate 2.0 not be necessary. If less late-age strength can be

3 Sodium nitrite 6.0 0.45 tolerated, at least to the degree that now happensCalcium nitrite 2.0 with concrete cast during hot weather, many of

4 Sodium nitrite 9.0 0.45 our everyday chemicals become potential anti-freeze admixtures. Chemicals such as automobileantifreeze, alcohol, deicing salt, and common

even available. For short-term needs, fertilizer can play an important role in protectingthere is another way to approach win-ter concreting. This report shows that Table 5. Everyday chemicals that allow concrete to gain at leastsimply adding chemicals to the mix 75%* strength after being at low temperature for 28 days followedwater will make the resulting mortar by room temperature for 28 days.(concrete) resistant to freezing. These Dosagechemicals, called antifreeze admix- by weight of Strength (%) relative to mortar

tures, serve to depress the freezing cement cured 28 days at 20'Cpoint of water and to accelerate the w/c (%) -5°C -70°c -20°Chydration rate of cement. Their effec- Deicerstiveness depends on temperature, Calcium chloride 0.41 3 121.3 94.9 80.0amount of chemical, and on how soon 6 133.2 130.2 110.3the structure is needed. 9 108.3 100.2 110.5Potassium acetate 0.41 4 76.0 77.5 78.0

In the situation where the tempera- 6 115.4 78.6 -ture falls below freezing but the work 8 109.9 90.9 -

must still get done in a hurry, there Calcium magnesium 0.41 4 87.7 - -

are chemicals that will work very acetate 6 96.2 - -

well. Table 4 shows four admixtures 8 97.9 75.8 96.4

that force mortars with internal tem- Water softenerPotassium Chloride 0.41 6 77.4 - -

peratures of -10°C* to gain strengths 9 85.3 79.5 -as if they were cured above 5°C. FertilizersMoreover, they gain more strength in Calcium nitrate 3 98.1 88.1 76.8the long run than control mortar 0.41 6 92.2 98.8 84.3

9 105.5 99.6 nacured at 200C (App. A). Thus, these Potash 0.41 6 90.8 - -

chemicals make cold concrete behave 9 86.6 78.9 -

as if it were warm. Sodium nitrate 0.41 6 109.4 - -

What happens to concrete if, soon 9 108.8 103.1 -

after it is placed, its temperature dips Car antifreezesbelow -10'C? All is not lost. Appen- Ethylene glycol 0.38 3 87.6 80.2 -

0.35 6 119.1 103.6 101.0dix A shows that concrete made with 0.32 9 95.8 93.0 88.6the Table 4 chemicals can avoid all Propylene glycol 0.38 3 90.8 83.8 81.9frost damage, even when cooled to 0.35 6 86.2 - -

-200 C for a long time. The concrete 0.32 9 88.1 80.5 -

may not gain much strength at this AlcoholsMethyl (wood) 0.38 3 75.7 - -

• It is important to know that this is a concrete 0.35 6 96.9 88.4 -

temperature. Because of thermal inertia and 0.32 9 79.3 83.7 -

heat generated by cement hydration, air tem- Ethyl (grain) 0.37 4.1 83.6 - -

peratures can be much lower-further study is Iso-propyl 0.35 6 90.0 -

needed to define how low. However, until addi- * The 75% requirement reflects what happens to concrete strength when it istional information becomes available, the cured at 50'C. In the table, blank spaces indicate that the chemical did not pro--10°C should be considered air temperature. duce enough strength; na means that tests were not conducted at that condition.

7

concrete from freezing when other chemicals are concrete from damage when the weather gets bit-not available. Although concrete made with these terly cold. Normal concrete is damaged as soon aschemicals may not develop much strength early the temperature drops a fraction of a degree belowon, Table 5 shows that the concrete recovers at freezing.least 75% of its potential strength once thawed. In For long-term applications, antifreeze admix-the situation where concrete can be placed and re- tures should be studied for their effect on theturned to when the weather turns warm, these freeze-thaw resistance of concrete. Preliminarychemicals should not be overlooked, data at CRRZEL and elsewhere suggest that the

The chemicals shown in Table 4 are used in a addition of certain chemicals into concrete can sig-wide variety of manufacturing processes, from nificantly increase its freeze-thaw resistance. Onepharmaceuticals to food additives and, as such, explanation for this improvement is that much ofare readily available most anywhere. Using them the antifreeze probably remains in the pore wateror the ones in Tables 3 and 5 in concrete is quite solution; it may not combine with the hydrationsimple. Weigh enough water for one batch of con- products. This, in turn, depresses the freezingcrete, dissolve the appropriate chemical or chemi- point of the concrete so that it experiences fewercals in that water, one chemical at a time, and then freeze-thaw cycles than would normal concretefollow standard concrete mixing procedures after exposed to the outside environment. Another pos-that. Once the concrete is placed, consolidated, sible mechanism is that the dissolved particlesand finished, cover it with a sheet of plastic for 3-7 somehow prevent the mixing water from fullydays to minimize evaporation. Because of differ- expanding, even when fully frozen. Laboratoryences in cements and how they react with various measurements of freezing mortars show that mor-chemicals, the field results could vary from those tars containing antifreeze produce less expansiveof this report. Thus, concrete mixes made with any pressure than do normal mortars. Less expansionchemical shown in this report should be tested translates into less damage. Thus, antifreeze admix-before use. It is recommended that trial batches of tures might have the residual benefit of making con-the concrete to be used in the field be made with crete more durable.the chemicals and evaluated for strength gain at Data also suggest that antifreeze chemicals canthe temperatures expected in the field. Until fur- increase the salt-scaling resistance of concretether work is done, it is recommended that a mini- pavements, bridges, and other structures subjectmum of 363 kg of Type I portland cement be used to deicing salts. Preliminary studies conductedper cubic meter of concrete, and that the w/c ratio elsewhere show that concrete made with an anti-should be no higher than those used in this report. freeze admixture is more resistant to freeze-thawThe aggregates should be ice-free. cycling in salt water than is normal concrete. Osmo-

sis is one explanation for salt scaling. Deicing saltsWHAT WE STILL DON'T KNOW create highly concentrated solutions on bridge

decks, for example. To equalize these concentra-Additional work should be done to make opti- tion differences, water may try to flow out of the

mal the combinations of chemicals used in this concrete, creating considerable pressures andstudy to improve concrete performance at lower causing the concrete surface to spall. Finding thetemperatures. The work should include not only right chemical dosage to reduce this pressure mayreformulating these chemicals and testing them translate into huge savings in reduced repair bills.for the best dosages, but it should include a search Current guidance forbids placement of freshfor other chemicals that can make concrete set up concrete onto a frozen substrate to prevent a hot-fast. Specific attention should be given to combin- tom layer of concrete from freezing. Preliminarying the Table 5 chemicals with others to enhance studies (Korhonen et al. 1997a) showed that con-their performance in concrete. crete can be placed on very cold crushed stone

An effort should also be made to determine the without problem. Guidance on when normal con-minimum chemical dosage required for protection crete can be placed on frozen substrates needs toagainst frost damage. Such a dosage would not be be developed, both to avoid any freezing and toexpected to promote much strength at low tem- determine how to thicken the concrete shouldperatures, but it might become routinely used in freezing be unavoidable. Work should also beall winter concreting as a vaccine against unex- done define the effect of various substrates on an-pected cold. The findings in this study suggest tifreeze concrete.that even small dosages of some chemicals protect Finally, the relationship between the tempera-

8

ture of the concrete and that of the air is not well 04.02. Philadelphia, Pennsylvania: American So-understood. It is a function of the heat evolved ciety for Testing and Materials.during curing, the air temperature, geometry, and ASTM (1997) C39-96, Standard Practice for com-other variables. Work needs to be done to enable pressive strength of cylindrical concrete speci-low-temperature concrete to be mathematically mens. Annual Book of ASTM Standards, Section 4,modeled. Much of the modeling already done for Vol. 04.02. Philadelphia, Pennsylvania: Americannormal concrete at temperatures above 5°C can be Society for Testing and Materials.used in this endeavor. Knowing how to predict Korhonen, C.J., and J.W. Brook (1996) Freezingconcrete performance is vital to efficient and safe temperature protection admixture for portlandwinter construction, cement concrete. USA Cold Regions Research

and Engineering Laboratory, Special Report 96-LITERATURE CITED 28.

Korhonen, C.J., B. Charest, and K. RomischASTM (1987) C305-82, Standard practice for me- (1997a) Developing new low-temperature admix-chanical mixing of hydraulic cement pastes and tures for concrete. USA Cold Regions Researchmortars of plastic consistency. Annual Book of and Engineering Laboratory, Special Report 97-9.ASTM Standards, Section 4, Vol. 04.01. Philadel- Korhonen, C.J., E.R. Cortez, T.A. Durning, andphia, Pennsylvania: American Society for Testing A.A. Jeknavorian (1997b) Antifreeze admixturesand Materials. for concrete. USA Cold Regions Research and Engi-ASTM (1993) C1231-93, Standard practice for use neering Laboratory, Special Report 97-26.of unbonded caps in determination of compres- U.S. Army (1998) Concrete and masonry. FM 5-sive strength of hardened concrete cylinders. 428. Headquarters, Department of the Army, Wash-Annual Book of ASTM Standards, Section 4, Vol. ington, D.C.

9

APPENDIX A: PERFORMANCE DATA

The following four tables present performance of the 28-day strength of control mortar cured atdata from mortar made with the chemicals evalu- 20'C. The majority of the strengths are given forated in this testing phase. They present the effect 7, 14, 28, and 56 days at 20, -5, -10, and -20'C.of the chemicals on the freezing point of the mix- Where applicable, the mortars are cured for theing water in the mortar and on the strength gain first 28 days at the temperature indicated aboveof the mortar cured at 20, -5, -10, and -20'C. each column, then an additional 28 days at roomFreezing points were determined by embedding temperature.thermocouples into cylinders of fresh mortar Table Al presents the results from chemicalsplaced into a -20'C room. From the resulting that had a propensity to cause increased mortarcooling curve, the freezing point was identified strengths at 7 days compared to that of controlas the point where the slope of the curve changed. mortar (see Table 1 in the body of the report forFor pure water, the slope of the cooling curve comparison) when cured at 20'C. Table A2 showsbecomes zero until all of the water turns into ice, the results from chemicals that had little tendencySince mix water is a solution, its solute concentra- to promote early strength or that may havetion increases as pure ice freezes out, which pro- caused delayed strength, but that provided freez-gressively lowers the freezing point. Thus, it is ing point depression merely by being in the mor-that the cooling curve slope only changes and tar. Table A3 lists the results from chemicals thatdoes not become zero as it does with water, The are liquid regardless of whether they act as accel-dosages at which the chemicals were evaluated erators or freezing point depressants. The Tableare presented as a percentage of the weight of A4 chemicals were combined in an attempt to cre-cement in the mortar and as molality-the num- ate an admixture providing both freezing-pointber of solute particles added per kilogram of mix depression and accelerated strength gain of mor-water. Because the formulation of all chemicals tar at temperatures below 00C. Time did notwas not known, the number of solute particles allow for the optimization of chemical combina-added to the mix water could not be provided in tions. This should be a topic of future study.all cases. The strengths are given as percentages

(z cc cz t-m co 1w mc t-c c oo c c)c) ( cc cc a cm~t M cc cd.m&mc c cc cc c o o oa c oc LI ncy) to

No t- -o -- r o Dt

u r * -:o c) oo oo oo cm -di m .c) c o~o o )c c )6 m~ c-S oo c - ,vc m~c c--coo = d z0 t- t - -- m- n a nm r o ý c c)t o t -m t -t

Lo c.. c m V o c'c - D ~c~-m cc'e) , .- (2c. ocwO -T-. ) a c)~ c-o m-- c) D" to mc t- t1o t

c10 cc l c~c ---:c ' 'R qcr 'R -- oq "R t- 0= q c N 0 f - cl N -- cR 0ý "O c cc c--0 c-c c lc -. c -co mc-l oc-c- cccc m-c m'cc co- oc-c ceO ccccc c"cc~cc - o c

(2C o omt "ccc r_-:o'3c ~~r~c-c -*c o- c2cc cccý.o oo- ED'

Nccccc-cl- tccc : cc rý cccv- tRc- cccc' ccoc 'tc tR tR 1- cic~' qc~' cccc cc-Pco mco - 0 c cc-c- cD )-cc C),w mcc- c co (zc- z . co .- -c co 00

0 cy)c, oc )00 c ) -0N r r:O rc.-.l-C. . -r rOr

0 r -~ c-.6 r c i c t 6qc wcý-c--cra; tcoccc 06 c r-c c i -4t4 cccrccr ccc-i -C5 i c-c ;: o c )

N c cc qc -c-c 00 0c-- ~Cr q c t cc c T cc cc N t- cc cc -- c- cc1 cc ccý 66 - o6ý -

t- -l -l -o -o Ccc c-c ) cs r e)" o ) t --cr1 c-c 1 -w

mc - -c) c-I c-c a)oc m0 mc- to c -c cc-0 c t- - cl-c c cDc 0co c) )cc j cu c- o - c-i-c cc-c c -cocc6 m o mc~ mccc cc cc cc coc m 4oi~ 4 co m c -c c5 6 -4 o cc c5c; cc

ccl

lp Nr-c c-c-c Ncc- crc,-- 0- ccý c-c- q-cc c-cc- N-c-c qc c--c-- cCc- -cc ýN o to -c-c c-c-c- c- mc- oc-cm c-c- m m-c-c m -c cD 00 0c-oo ccocc2 o ~cccc Oo c)

N cc cc c-o) ~ - 0-cc cccc cc cc c co- c cc l - Iv~- cc) cc cc) 66 6 66

co ~ cc c-i c-cco cc- c-c.c :-cc c~ o- ec-cr o ro m o ýD - c--c cc

cc1

cocccm c-c6occo6 o6o cco C 4 o- ico6mo: 2a 22( ccM mc- -0c cl- o) c-cc--c o a)c- coo

cc cccc- ccc"-c0c - cc l--: q c-c N -- c. ccccc cc cc cc- - ýcýcýomcac o cc c) -iýcm .i c -o C)C o)mco *cD a c c Mco cD c)cDc cD oc c--V) rcc- C cccc ('cc aO c-.cc ..C c .

N oýo' --- c-cc-c c-c0 - cc t-" - c )- cc 0O ) 000 -cc c- c) m (n do a) o c-ccccc t"i r i6- i56 " C 6c 4" ici6v -- m Ccc -o i- cc c --cc ccc-- cicccc

0 DClo cc- Rcq ~cc-cl -c- o c-ý c t t-1-: '!cý c-- , ý ý N--c coqc q *: cl(-Ic-Ic-No 6 4 -cc cc cc 4.4 t-c c 0 t0 c-.c cc cc cc , 666t cc cc 666 m c cc cc cc cc~-~ c5 000oIz C c ) o ca -CCC CC) CC

o q Ncc- c-sc- cc -i c t: cl crc'.I t-crc 00- 0c-ccn ýLýt t-cc ý L cc--c c-ccc Oq cc- I-,c-c- N Ocr-c0 e ~coc- cq-cc-Z c-ccccc cD -c cz cccc -c-cc- m (D CV) -0) w m co t--c -c-cc-c o

II a I II I I

mcc~. -O oo 1w 0czc-c cu cw m mc-- -c--cm o -o-c- c- o o cco m -o- c'cc ~ C- cc-o~ ~ oR ~ - r c r.cac a--rc- c-c- "-~-sc ~C-- c-cc- c cc-c tcc-c -c to-c mc c 15cc

~ ~ cd ocsc. Cl c -w6O c-ic D C- i - i 6.~- vi -c 6cýi -si 6-~c- 6 c;cc c-icc

LI 0 000 000 6 00 cool5-c) Cc)c 0 cc a ~ ccccc- cccc 6c 6c 6666cc ; c---c cc c-c cc-cc6

ccz

cCL

E E

CL cu c cc cm~ CCc c - -o Q 0- co-

cc c E) c u cc v) C c

12

cam o c c a~- t, m cm cea D o ý; - t t - co m m' co coa coa co a6D m cma co ~-Iz m cz co w t, c)a a) co m coNj lz to toD DN t- r.)C~ a r- =Ca ýo t- -- co .: oo~ a)t o

cýc

ti :r m- ).r c ' -u eI o m t-mt oc o m w wi ;

to

No cl o -- ýa l Nlý 0 -- ! -1 u-t:a ýc , Rt 1q1 ýo l l qwG , ýI!c qc ý

m cDCCO qrC a c)) 000v (n) 0000a) m ca (Dc ca oo Ca ta2 ca ca m a a c ~ 6 ca 2! ca-la ma to ca 6 a) ca)

2) 2 to~CC m' -- C2N -t

00C m m ID ý- m - -- t- C C

m co " a ca a)~-11 oo ma c o 0c) NJ Ca)a ca m) w) Ca a) C C, -W ma~C)C M a C>00 Ca c a cD

w )c mm =; MC? 0ý cc- ýo 2a -'n t ,mwL -a)2 , q 0 ot. omt *

C.C a). R- lN q IRc lL ýc q N N IN w- ý cý- o l ýc ,:tui o6- cici m - " c o t o " t-c , o " t z Dw t io c oc ,

lo c \ D c)m I l y o - c ot ~ )o -o -c

c- N cD0 Cf- r00 C OC r0 000 r. 000r:

m J m - to to ma J - m a to oa 0o 0m 0 0 0 ,- 0 l cD ac 0 ý "~ cl 0 ý a) a) a) ý Ca ý qa oa 0l 0ý 0ý ta Ca: La C aC Cao 0 0 a a

a) o Lr ýr-c --z C ou)c- t- c cla t-- Lnt - wc') -D Lo CD -n NJ .v m- L- -Rj -o

N)j co mD I 0 D JNJ0 D- Nr C) N 00 C-C -C)Dt N )C ) -0 CC NJ00N 0 C) -CD CD-0

CCDNr r:N W C~NN C C5 aC zr.r rr.C

m. 0 0 0"N ) > -, -)m o o Z N -.r0 0mm =C 0 ! q(R- "0 c6 6 j 0 0 6 .6 . 6 6 7; 0 6 6 cz-N5 D 6ci c= 0 0c)Na 0c.co m cD c

on mV: cl t- ýc 1 m m-z Dm c l ý', a l ýR C ! t D w1ýmc Icýc Rc!o

.d I o I ý C) ID Iý m Il m to Il w )m t

44 C D to C 00Nc m cy)D NJ D J - m m 'r- )CD c l)~ .Zr " )

CCCC

w coLfolac , ' ' o-f-lm 'r. r. vi 6 4 too o6 4ooo,;t-- 4L6 -o6 c ~ ~ 66666-666I =6=6I

42 z oa , 5) 5t -- ca-vc oc oc) c C oclc oo v c a)

C) r~ a) C: a r

c ) o) a) a) 6 .C )c Dc )0666c;c ;666

a)a) C. )c)~ ~ ~ ~ w~ m E oo w o - Do ommI rý 2c owmý om..ý -vww t

a) -u au

E -0 cu E

-a -!2 2 ca R -C 0 f c.C W) -- ~ 0 r) Q)

13

NJ (Ct~- t t-- a r D wI omý D L D m I

-: t- r qq qN ý -0 ý a - !c : c qC

'1'C oMm o- n t.omt o)Dl z cm oc omc t-C0 ( 0 0( C)

tý~~0 ( ( ooo oý m m = w m c c Dt w = 1,"r

Ic l icýN - ýc )I l p cý ýI p u ýIREl ýcc ýt

0i 000 00 ) 0)00000 0000c> 000

41 c" R -NJ "ý ý ;;-1)--0)1l -c)- , o-z

C.-a

o 0) (cc0 m) t-. m"( ) 0- CD .--CV) t- C.. C)0 0D m~ - m~J to J m 00

oU -,I - - ---- -ot r o t mc]- C) m c m ) M - cpC.J ~ ~ )o - - 4 t 64 4 c 6t:c ~io v ~ ~ v ; o

2 ý t, 00 cW c o a)o o ot t- t, t D 't c

0qI ý C ýqc ýc fLIcDc -ývc )mmmv c )a )c DC )c Dc )c Dc

c1 C mm) DC . o ooc)cI o - oc4 000 c5C0 o6-00005c0Lic oc c 5ci 0ci ci

0 - .o .o .o c : om o wt D ( rNJi c6~0 0)0) 6 c Li ; D000v0000vi 000006 00: ý c

000 qm m p NcýV!e ýL

w 0)t-4 ;ý m m cct o " ;IDI ( DU =ýo mI o c)" t

1- i( I c R)o 1 R LýN q l ý o

Co c) cD c C) 0) C.. ) (c) C). C) c) cC c.') cDC') C) c>C) 0) C') oC C)C)00 C') C

I ~ ~ ~ ~ ~ E o -r- IIIIII

a0 o

m- ~ on co r o oc0s m 2~ CC)~ C 0 o

14

coCc -i V! tqu !u ýc ýv l I i L ýc ýN c Rc q

w cco Lýcsw c ~ c;ccLr c6~c~ ccý r-ý y 6 ~ - c 4 6

Lr ~ l 6v 6tic 4o v - TC: C a- a CCcn - a)(='cDco m m --oocemc); o-c 2( t o aocC2 (ncctcco- - oo t- a)

to

ci1 (cc cc cc ccoccN t,- --: -cn1ýIqr-o RI IV -c C clcl Cý = t-: -c Ncl 0 qa a m oo CC cD cn eoo-.- cc, m LC-- O c- a)O cDO

Q)c ! i rl iI jt-o q pU qq cýN - ý 17 c l qc c- N l- t- p

m oc ) oc CO OC~) (DD O C O 0 co D C -- C DC t, cccc 'dcC, cc m m

N n -" -l co co -om ýoI -r " - (--C)m t o - - m

ccl]

1ý -z 'c c1, - m cD m 1vtM CO ~ ~ OD CD- O co- C1 m mm c)t CO l

co - -- cc2 ot L-6 - t L-- ~i cc 6 c i Lm - CDt Dt-t--z

to m - CC mt'J ~ t ,m o m G o(a acc~o o-C- o-- oC-12o2 -c

t-- t- w 6 m CiO- mC C'i -:C-)CC-i -' C(*)"C'i 6 - 46 6 C ý-

ID -w " c) CY )

0l iC Nf~ V! cc-c---CO ec- ~OC-- N io l ýC IR IR o ' i

N cc) t- ID0 C w 0 - -.d -- = " a)wC - D - m;:t mmw - ccccot-t DC ý D ý ýr cc mCc---c-) ~ c)C)~ 0)~ c ~ C -oclc ~ c-o

S .pm IRc K , lt- ýC - qL D0 KnN 0m t~(- ~ f- w O-Om'Cl) m 0 C0 -- a c'. m '-0 0-Cm 0Ct- " 0 .d OOO)

0lc Cý 0)c C' ) C-)wOC) m m C\! r- OCDc0"! t- CO CD M 00 -)oo e 0Oco 2C~ 00 2~fC~ -~J cy) cC)CCM)--c-..co- C

00 t- v -V- -c~ Cl)o M CD DoO r(O )cm D m )CC.((Cc) c I) o Cl) m M --d CD mJCl--Zt C)t.mv) o 6 6 C-- ccc) 6o c o c) 0( ;cei 4 ---c -,:vcccC -- :C6 a 0-4 " .*;C6 L6cc- C(Ccc;06C-c- .

C-- ( CC ýD LO -Ir " f CYC c l l ~ l e l l l - Cl) " -.(

ca-toCC) 00 CD Co CO oC13CM c'JC'- cC) C- m 0"c 1 -d ~ Wt00 Mt0-t -I 4= mi(7 cc6 M" cc-C0C)C.664o . 6646Di0 0C)CDO 0CC ;ccccO t-r)i6

cc C6 cccc ic -Wc~c cý- C4.OO~)cc C -v -w 4 c-r-j-c6 o6 o6~.

e-C.c cc C c c cc m c (o Itce ce; C C C C

m w m w m 00 a)C) to t-CC CC) 0C) t- C

C)O c cC) oC DCC ) cD cc) c> = a) = cýcc) Slc) cDmCD D C) ( c D c) CC) CD(C CD = ~c D c)

v.- -I C)

-2.C c a )t Wc C) a) Co

N.2

N.4~~ ~ 0. C- co~ 02 E -C=c Mccc rd -Ccc mu.0( (1c )c ~

:s) _Io Cl', V U V V ccc V 0

J45