The Ultimate Geotechnical Engineering Challenge

Some of the Problems and Their Solutions

It would take pages to go into any detail about solutions to the problems Les Karafiath and I had to address, but here are a few of the geotechnical challenges we faced and how we managed to resolve them. A question and answer format is used here for this purpose.

Q. How do you determine the near-surface geologic profile of the Moon, in particular, what is the thickness of the “dust” on the lunar surface?

A. Fortunately, the research section that Les and I were in was called the “Geo-Astrophysics Section.” As indicated previously, most of the other members of the group were scientists. Some of them were planetary scientists, so they gave us some clues as to the geology of the Moon, but nobody knew for sure. One of the most controversial topics at the time was the composition and thickness of the lunar regolith, i.e., the surface layer or “dust” layer. The controversy stemmed from the fact that the weathering processes on the Moon are very different from those on earth. For one thing there, all of the scientific evidence at the time suggested that there was no free or adsorbed water on the Moon. So the products of chemical weathering of parent materials on the Moon as we know them here on Earth would not exist. Consequently we could dismiss the presence of clay minerals on the Moon, but not necessarily clay-size particles. In addition, there is no atmosphere on the Moon, so the surface of the Moon gets hit by billions of meteorites and micrometeorites each year. That is not the case on Earth where most meteorites burn up in the earth’s atmosphere before the reach the surface. Thus physical weathering by comminution due to meteorite impact plays a major role in the physical weathering of rocks on the Moon, whereas it is virtually no factor here on Earth.

Weathering due to thermal expansion and contraction can be a factor at some locations on Earth where there are significant diurnal and/or seasonal temperature changes. On the other hand, thermal weathering plays a very important role in the formation of lunar soils. In the day, the temperature on the Moon can be as great as 123º C (254º F). At night it can cool to as little as -233º C (-387º F). That’s a swing of over 356º C (640º F) in a 24-hour period.

Other factors such as electrostatic agitation of fine dust by charged-particle bombardment from the solar wind are analogous to mass wasting here on Earth, especially in view of the Moon’s gravity being 1/6 of that on Earth

These and other factors led to much speculation on the part of scientists as to the thickness of the lunar regolith. I can recall attending a conference in 1966 where Dr. Thomas Gold, a renowned but rather eccentric astrophysicist from Cornell University, suggested that the lunar surface was coated with a deep layer of fine rock powder and warned that astronauts and landers would sink out of sight. Others suggested that the lunar dust was only a few inches thick. These wide variations of opinion were disconcerting to me and Les since we were engaged in an engineering design and did not have the luxury of academic theorizing. The Lunar Orbiter program (1966-67) provided high resolution photography that could be used to obtain good estimates of the thickness of the dust at specific locations on the Moon, but not necessarily at the proposed LM landing sites. Scientists used photos from the Orbiter program to estimate the thickness of the dust layer from the depth of tracks left by boulders that had rolled down crater walls, such as the one shown in Figure 2. Les and I could be sure that at least at those sites the dust was not so thick that the boulder disappeared. By the time the Surveyor Program was in full swing (1966-1968) to verify the thickness of the lunar dust at prospective LM landing sites, Les and I were well on our way to providing geotechnical input to LM project engineers for the design of the LM footpads. It was satisfying to us that much of the data received from the soft-landing Surveyor spacecraft confirmed our initial estimates of pertinent soil properties.

Q. How did you identify the type of soil at the surface and estimate its engineering properties such as gradation, density, and shear strength parameters without physically testing the soil? No guesses allowed – the stakes are too high.

A. The planetary scientists were convinced that the surface and near-surface geology consisted mostly of basaltic igneous rocks. Therefore, because of the type of weathering processes on the Moon, the lunar soil was most likely a basalt powder. Before the results of the Surveyor program were made public, the grain size distribution of the lunar soil could only be surmised from comminution theory. There were some excellent scientists at Grumman who used comminution theory to estimate the grain size distribution that Les and I could expect. Their estimate turned out to be remarkably close to the distribution obtained from tests on actual lunar soil samples returned to Earth by the Apollo 11 astronauts. Based on their input Les and I comminuted basalt rocks to prepare a material with the grain size distribution of a silty fine sand. We used this material in all of our experiments as a simulated lunar soil. We hoped to perform experiments on the simulated lunar soil in a simulated lunar environment in order to determine values for the parameters needed to perform geotechnical analyses and design. For example, we needed to estimate the shear strength parameters of the lunar soil (c and φ) in order to perform bearing capacity analyses of the LM footpads and to estimate settlements. We also needed to make estimates of the in situ density (ρ) or unit weight (γ).

Figure 2

Lunar Orbiter 5 image of a boulder that rolled down a slope and left a trail on the Moon’s surface within the crater Vitello. Date/Time (UT): 1967-08-17/04:04:40; Distance (km): 172. (NASA photo)

Our laboratory experimental program was unique, but, as we found out after the fact, the results were not very reliable. For starters, the Moon has virtually no atmosphere, i.e., the atmospheric pressure is equivalent to a “hard vacuum” i.e., atmospheric pressure <10-12 mm of mercury. Therefore, if we wanted to determine the effect of a hard vacuum on the engineering properties of the simulated soil, we would have to construct a vacuum chamber that could be pumped down to that pressure. With the technology at that time it was virtually impossible to construct a vacuum chamber here on earth that could be pumped down to 10-12 mm of mercury (1 mm of mercury = 1 torr). Going from 10-9 to 10-10 torr required an entirely new technology that even the most sophisticated laboratories in the country did not possess. After a few months of their own research, the outstanding technicians at Grumman developed the necessary technology, only to discover that at about 10-10 torr the walls of our stainless steel vacuum chamber began to out-gas to the extent that the chamber pressure became virtually constant, i.e., the gas molecules naturally existing in the materials from which the vacuum chamber was made began to be sucked out of those materials and into the chamber due to the pressure gradient. In fact, at about 10-10 torr we really didn’t know whether we were actually measuring that level of vacuum since the ceramic tip on the pressure sensor was probably out-gassing also. Figure 3 is a picture of me in the Soils Laboratory at Grumman looking rather puzzled at some pressure measurements.

Figure 3

A photo of the author in the Soils Laboratory at the Grumman Aerospace Corporation, Bethpage, Long Island, New York (circa 1967).

Strength Testing

Figure 4 is a picture of the stainless steel, quadruple-mold specimen holder that Les and I used to pre-stress simulated lunar soils for exposure to an ultrahigh vacuum, i.e., pressure < 10-9 torr. Unfortunately, we could not test the specimens in the ultrahigh vacuum chamber for a number of reasons, but mainly because the necessary feed-throughs invariably leaked. Therefore we exposed pre-compressed specimens to an ultrahigh vacuum for a period of at least 24 hours and then tested them in unconfined compression after their removal from the pressure chamber. The observed increase in unconfined compressive strength of the specimens subjected to the ultrahigh vacuum as compared to pre-compressed specimens in air suggested that bonding between molecularly clean surfaces (i.e., no water or gas molecules on the surfaces) had occurred as a result of a complex interaction of thermal and mechanical stresses in the ultrahigh vacuum environment. Unfortunately, we could not extrapolate the results of those tests to lunar conditions with any degree of confidence. Therefore we had to seek another way to estimate the shear strength properties of the lunar soil. That is discussed as part of the next question.

Unit Weight

There was a relative wide range of estimates of the mass density of the lunar dust available in the scientific literature. I did not feel comfortable with any of the published values mainly because I didn’t entirely understand the science behind the ways they were determined. I turned to Walter Egan, one of my colleagues in the Research Department for help. Walt’s background as a physicist was in optical sciences with years of experience in photometric and polarimetric measurements, of which I knew nothing except what I had read about in his internal Grumman reports. Walt was interested in our problem and the two of us conducted research on the use of spectral photometry and polarimetry to establish a relationship between surface porosity and polarization for the simulated lunar soils Les and I were using in our strength experiments.

Figure 4

Quadruple mold, stainless steel specimen holder used for testing simulated lunar soils subjected to an ultrahigh vacuum environment

We used the relationship derived from these experiments to determine a range of porosities for the lunar surface material based on available polarimetric signatures of the Moon. By calculating void ratios from porosity and by assuming the specific gravity of the lunar soil particles to be 2.7, we were able to calculate equivalent unit weights for lunar soils and adjust them for the reduced gravity. The range of our computed values fell within the range of values found in the scientific literature. Although I felt more comfortable now, there was still no sure way of knowing how good (or bad) our estimates were until the results of some of the Surveyor experiments had been analyzed. Incidentally, the results of experiments performed on some of the lunar material returned by the Apollo 11 astronauts showed the specific gravity of the solids to be 3.1. At least we erred on the side of conservatism when we assumed 2.7.