An Experimental Study of Strength of Young Sea Ice

Values for the strength of sea ice are extremely sparse in the literature, and available values show a wide, unexplained variation. Results of in-place cantilever beam tests presented in this paper show a definite relationship between fl,exural strength and brine content. Values for Young's modulus are also experimentally determined. It is shown that the bearing capacity of a sea-ice sheet is dependent on the brine content and that thin ice sheets are capable of supporting a large 'super load' beyond the force necessary to form the first crack. Introduction-Detailed studies of the basic physical properties of sea ice are very few. Malmgren's [1927] work still remains the standard reference. In particular, information regarding the strength characteristics of sea ice is noticeably absent. Inasmuch as increased activity in the polar regions during IGY has necessitated the use of unprepared sea-ice surfaces as floating platforms for scientific activity, the bearing strength of sea ice has become an extremely critical problem. The specific objectives of the present study are: (1) to determine the relations between the flexural strength and Young's modulus of sea ice and such parameters as ice temperature, salinity, and density; (2) to calculate the bearing capacity of young sea ice; and (3) to develop theoretical relations for the analysis of sea-ice strength based on the observed distribution of brine and air pockets so that it will be possible to extend experimentally determined strength curves. The present study considers only in-place cantilever tests on thin ( <40 cm), newly formed sea ice. This material was chosen since there is a large variation in its brine content and it is thin enough so that a large number of strength tests could be performed in a reasonable amount of time. This was considered necessary since previous work by Petrov [1955] has shown that the results of small sample strength tests on sea ice show a large scatter. Also, these tests duplicate conditions on the warm underside of the ice where the ice initially fails. The sites of the field tests were North Star Bugt, Thule, Greenland, and Hopedale, Labrador. The· tests were performed in areas of homogeneous sheet ice. Testing started as soon as the ice became strong enough to support the weight of the test equipment. Tests were made on this ice sheet at regular intervals during its growth up to a maximum thickness of 40.2 cm. This procedure, however, made it possible to collect only a very small 641 amount of information during the freeze-up period since at any given ice thickness the ice can be tested at only one temperature and salinity. To avoid this difficulty, a number of large ponds (rectangular ice-free areas) were cut in the ice. These· ponds were allowed to freeze on different dates so that ice of varying thickness and salt content could be tested during a period of a given air temperature. The majority of published sea-ice strength tests have been made using small samples that were removed from the ice sheet and allowed to reach equilibrium with the air temperatute [Butkovich, 1956; Petrov, 1955; Veinberg, 1940]. This type of testing is not suitable for studying very young sea ice, because when the specimen is removed from the water (1) a large portion of its brine content is lost by drainage; (2) it is often· not capable of supporting its own weight; and (3) it is impossible to re-establish the exact in-situ temperature gradient. Also, the results of in-place, cantilever tests show a much smaller scatter than most small-sample tests [Butkovich, 1956]. Therefore all the tests were performed in place in the ice sheet so that the physical properties of the specimen would be identical to those of the over-all sheet. Flexural strength-The general technique for obtaining the in-place flexural strength of sea ice was used by Neronov [1946] and other Russians, earlier, to measure the flexural strength of lake ice. First a U-shaped channel is cut in the ice sheet with an ice-pond saw. This channel isolates an in-place cantilever ice beam with one end attached to the sheet. Both pull-up and push-down tests were performed on these beams. On the pull-up tests a chain is placed around the end of the ice beam and attached to the end of a wooden lever. A force is then applied by rolling a heavy barrel from the fulcrum to the end of the lever. From the weights and positions of the chains, lever, and barrel at the time the ice beam breaks, the force 642 WEEKS AND ANDERSON applied to the end of the in-place beam can readily be computed. When the ice was very thin, the force was applied manually, using a dynamometer to measure the force at the end of the ice beam. The push-down tests were performed using an additional lever that was anchored to the ice sheet. This lever transmitted a downward force to the end of the ice beam. When push-down tests were made on thin ice, a known weight attached to a dynamometer was lowered manually onto the end of the ice beam. In this case the force at the time of failure is simply the known weight minus the dynamometer reading. The cantilever beam width was kept roughly ! to 2 times the ice thickness and the length between 3 and 6 times the ice thickness. The average loading rate during the tests was 0.5 kg/cm2 sec. After failure the broken segment of the ice beam was immediately removed from the water and three measurements each were made of its length and its height and width at the failure cross section. Three measurements were necessary because of the irregularity of the break. A 7.6-cm diameter vertical core was taken from the sheet in the area of the tests, for the determination of a salinity profile. The ice samples from the core were immediately TABLE 1 Results of in-place cantilever beam tests Type of test Ice Ice type Test U = pull No. of S Stan. Aver. ice Skel. series Date up tests <f dev. <Tk <rm Av. sal. temp. thicklayer S = sheet D =push ness P = pond down -----------------kg/cm' kg/cm• kg/cm' %o •c cm cm 1 Oct. 29, 1956 u 2 0. 72 0.14 1 . .19 1.40 10.7 -2.4 12.7 3.6 s 2 Oct. 31, 1956 u 6 1. 70 0.15 2.45 2.75 10.4 -4.2 16.8 3.6 s 3 Nov. 1, 1956 u 7 1.50 0.19 2.07 2.24 8.8 -3.0 18.8 3.3 s 4 Nov. 2, 1956 u 5 1.25 0.16 1.68 2.10 8.9 -3.6 20.1 5.0 s 5 Nov. 3, 1956 u 7 1.36 0.06 1.82 2.11 8.3 -2.4 20.5 4.0 s 6 Nov. 5, 1956 u 3 1.05 0.07 1.34 1.47 8.1 -3.4 23.6 3.6 s 7 Nov. 6, 1956 u 3 1.15 0.10 1.47 1.63 8.0 -4.7 24.1 3.8 s 8 Nov. 6, 1956 u 7 0.86 0.04 2.16 1.91 13.0 -3.5 7.6 2.5 p 9 Nov. 8, 1956 u 7 0.92 0.08 1.66 1.62 12.3 -3.0 10.9 2.7 p 10 Nov. 8, 1956 u 2 1.11 0.28 1.66 1.82 11.0 -4.2 15.5 3.4 p 11 Nov. 9, 1956 u 5 1.32 0.16 1.80 1.95 10.5 -3.2 16. 7 3.0 p 12 Nov. 10, 1956 u 6 2.49 0.30 3.00 3.20 7.4 -5.0 31.6 3.7 . s 13• Nov. 12, 1956 u 11 3.03 0.29 3.49 3.93 7.5 -5.8 35.4 4.6 s 14• Nov. 16, 1956 u 6 2.48 0.15 2.87 3.06 6.0 -4.2 40.2 4.0 s 15 Nov. 16, 1956 u 7 0.63 0.14 2.06 1.66 15.2 -3.0 6.3 2.4 p 16 Nov. 17, 1956 u 8 1.39 0.24 2.04 2.04 9.6 -4.4 15.7 2.8 p 17 Nov. 20, 1956 u 5 1.88 0.30 2.54 2.69 10.1 -6.4 20.4 3.3 p 18 Nov. 20, 1956 u 4 1.38 0.06 1.92 2.02 12.4 -6.4 18.1 3.2 p 19 Nov. 21, 1956 u 7 0.98 0.15 2.09 2.11 16.0 -6.7 8.9 2.9 p 20 Dec. 24, 1955 u 4 1.29 0.28 2.13 . . . 11.8 -3.6 12.7 ... s 21 Dec. 25, 1955 u 3 1.30 0.30 1.85 . . . 12.4 -3.2 17.8 ... s 22 Dec. 28, 1955 u 12 1. 72 0.17 2 18 . . . 9.3 -3.8 16.5 ... s 23 Jan. 3, 1956 u 6 2.15 0.30 2.63 . . . 8.3 -4.8 28.9 ... s 24 Jan. 4, 1956 u 6 2.47 0.36 2.99 . . . 8.4 -5.5 30.5 ... s 25 Jan. 11, 1956 u 5 1.41 0.20 1.67 .. . 7.4 -2.7 36.8 ... s 26 Oct. 31, 1956 D 5 1.95 0.62 3.20 3.31 10.4 -4.2 17.2 3.1 p 27 Nov. 1, 1956 D 5 1. 73 0.63 2.33 2.61 10.0 -3.0 18.5 3.5 p 28 Nov. 2, 1956 D 4 1.27 0.35 1. 72 2.17 8.9 -2.8 20.0 5.2 p 29 Nov. 3, 1956 D 7 1.39 0.17 1.83 2.10 8.3 -2.4. 20.8 3.7 p 30 Nov. 5, 1956 D 3 2.74 0.12 3.54 3.86 7.6 -3.6 23.4 3.7 p 31 Nov. 6, 1956 D 3 2.33 0.59 2.98 3.25 8.0 -4.6 24.1 3.7 p 32 Nov. 6, 1956 D 4 1.25 0.11 3.08 2.75 13.0 -3.0 7.7 2.5 p 33 Nov. 8, 1956 D 7 0.84 0.21 1.49 1.33 12.3 -3.0 11.3 2.3 p 34 Nov. 9, 1956 D 4 1.41 0.14 2.02 2.01 10.0 -3.9 17.2 2.8 p 35 Nov. 16, 1956 D 3 0.50 0.19 1.63 1.32 15.2 -4.0 6.3 2.4 p 36 Nov. 22, 1956 D 5 0.93 0.19 2.00 2.04 17.6 -7.0 8.7 2.8 p 37 Nov. 24,-1956 D 7 0.96 0.30 1.48 1.63 11.3 -4.8 14.3 3.3 p 38 Nov. 26, 1956 D 7 0.91 0.11 1.50 1. 71 15.2 -6.0 12.8 3.4 p • These tests were performed by Lyle Hansen of SIPRE. EXPERIMENTAL STUDY OF SEA ICE STRENGTH 643 placed in tightly sealed glass containers and allowed to melt. The density of the resulting solution was measured, using hydrometer floats from a salinity kitl manufactured ·by G. M. Manufacturing Co., New .York 12, N. Y. All values were corrected to 15°C, using a table supplied by the manufacturer. The tempe·rature of the beam was measured at two or three levels using Weston dial thermometers inserted into drilled holes. The average temperature of the ice sheet was then determined by using both the results of the dial thermometer measurements and temperatures .measured by a series of thermocouples frozen into the ice sheet at 10-cm vertical intervals. Typical temperature and salinity profiles during the period of testing are shown by Anderson and Weeks [1958, Fig 8). The flexural strength is computed from