OAKLAND -- The thunder of falling concrete continues here as demolition teams dismantle the concrete-and-steel leviathan that collapsed in an earthquake Oct. 17. The scene contrasts sharply with that of June 11, 1957, when double-decked Interstate 880 loomed as a bright example of human ingenuity. On that day, the leading lights of the nation's most dynamic state highway building program joined Miss Oakland, Sharon Gleason, in cutting a ribbon to open the $10.2 million "Cypress Street Viaduct" of the Nimitz Freeway. Three years in construction, the roadway would lift west Oakland's traffic problems above a neighborhood congested with industrial facilities. The 1.8-mile roadway's two-tier design, then unique in California, and the mass of concrete and 29 million pounds of reinforcing steel required complex design calculations in an era before highway engineers used computers. Today, engineers and geologists say the fate of the viaduct and the 39 known victims of its collapse was sealed somewhere in those calculations and in the raw seismic power of the shifting earth. While a more complete explanation awaits the findings of an investigative panel appointed by Gov. George Deukmejian (R), a consensus is emerging that the failure of key joints in the mammoth structure triggered the collapse. "The clear cause of the failure was at the hinge joints, where the {first-level and second-level} columns meet," said Gregory Fenves, a civil engineering professor affiliated with the Earthquake Engineering Research Center at the University of California at Berkeley. Interviews with engineers indicate that, long before the collapse, the Nimitz was known for engineering defects. But it remains uncertain whether the California Department of Transportation (Caltrans) was told of them. Caltrans officials insist that they were not. J. David Rogers, a private engineering consultant in the Bay area, said the designers failed to engineer the roadway against "creep," the tendency of massive concrete structures to sag between supports. Rogers said his professors used the viaduct as a case study of creep when he was a graduate student at UC-Berkeley in the mid-1970s. "If you drove down there at 38 miles an hour in an ordinary General Motors car, it was like a roller coaster," he said. "It was a wild ride." Caltrans designers were able to compensate somewhat for sag by repaving, but Rogers said a more ominous flaw came to the attention of the late Jerome Raphael, an engineering professor at UC-Berkeley. Raphael, who sent Rogers to make a preliminary analysis of the roadway in 1978, "conceptually recognized" that the double-decker was "so heavy and so high off the ground it was going to sway a lot in a quake," Rogers said. Further, he said, Raphael believed that the weakest elements were the upper columns, which were connected to the rest of the structure by hinge joints. In the early 1950s, when the Nimitz span was designed, elevated highways were not engineered to withstand lateral forces released by quakes, several engineers said last week. Instead, designers were concerned only with downward forces imposed by the structure and traffic. Raphael recognized that quake-driven lateral forces would cause the upper half of the Nimitz structure to rock and possibly topple, Rogers said, but apparently never pressed the point with Caltrans officials. "I asked Raphael once, 'What is going to happen to the {Nimitz study}?' " Rogers recalled. "Are we going to do anything with that? He just said, 'There is no money for multiple-column freeway structures.' It was not like he went to the governor's office or anything. He was a very reserved gentleman." No money was available to study multicolumn structures, state officials said, in part because the San Fernando quake of 1971 toppled a single-column freeway interchange, creating pressure to retrofit that type of elevated road. The state's multicolumn structures, meanwhile, survived that temblor and so, in the hard arithmetic of budget restraints, nothing was done to buttress them. "I think {the Nimitz} was recognized as a potentially hazardous structure, but there were other hazardous structures," said Wilfred Iwan of the California Institute of Technology. "Where do you start? That's the question." Then came the Loma Prieta quake, a temblor registering a magnitude of 7.1 on the open-ended Richter scale that knocked Caltrans's priorities to the ground Oct. 17. Now investigators must determine whether fatal flaws were present in the design by Caltrans engineers or in the construction by the now-defunct consortium of Grove Shepherd Wilson & Kruge of California Inc. Engineers studying the calamity have offered several preliminary explanations for the collapse, and all emphasize design, not construction, elements. Among them: The hinge joints, also called pin joints. These connections were used to add needed flexibility to the structure but were vulnerable to lateral forces. Steel reinforcing at these joints was limited, several engineers said, allowing them to break under stress. Horizontal bracing. Jim See, a field engineer with Caltrans when the Nimitz was designed and built, said the structure lacked sufficient horizontal reinforcing metal installed to complement the many vertical rods in its columns. Engineers at the UC-Berkeley quake center, as well as Iwan and Rogers, said they believe that the horizontal bracing may have been a factor but was not as significant as the hinge joints. Soft soil. At the Nimitz's north end, landfill soils are suspected of contributing to forces that toppled the structure. Rogers said his studies in the 1970s suggested that the unusual length of the elevated structure and its relative straightness would leave it vulnerable to longitudinal wave forces, a phenomenon that some engineers say combined with soft soils underneath to undermine the roadway. The investigation is expected to resolve these technical questions, many observers here said, but allocating blame may be difficult. "We have to realize that the design team's knowledge in those days and awareness of earthquakes was much lower," Rogers said. "Had they had the knowledge, the tools to work with did not exist." WHAT HAPPENED ON I-880? Engineers studying the 1 1/4-mile section of Interstate 880 that collapsed in Oakland think the key may lie in the "hinge joints," vital segments of the concrete columns that held the levels of the double-decker freeway in place. 1) "Hinge joints" were included in the freeway's column design to absorb weight and vertical movement caused by traffic. But during the Loma Prieta quake the ground also shook laterally, creating stress the hinge joints could not withstand. 2) Few steel reinforcing rods snaked up through the joints. The upper and lower columns are honeycombed with these "rebars," but the hinge joints contain only a few. 3) During the quake, the columns acted like weak legs on a wobbly table. The lateral movement caused them to snap outward at the weakest point -- the hinge joints -- freeing the upper deck to pancake down onto the lower one.