Imágenes de páginas


on the cliff. It would manufacture hydrogen gas from methane shipped from the continent in the liquid state. By-product carbon monoxide and additional methane would be burned to power gas turbines to drive the compressors, which would be used in the liquifaction of the hydrogen, as well as of oxygen from the air. Some liquid nitrogen would be produced for the cryogenic stretch-forming of rocket propellant tanks at the assembly plant. At a mixture ratio of 5.5, the cost of hydrogen and oxygen, including taxes, plant, amortization, profits, and propellant losses, is expected to be 5 cents per pound, in the rockets.

A minimum of fabrication would be done at the assembly plant. The larger propellant tanks would be welded from sheet metal, and cryogenically stretch-formed at the assembly plant. Smaller tanks would be purchased on the mainland, and cryogenically stretch-formed at the assembly plant. Practically all other components would be purchased already assembled, and transported to the assembly plant. It might even be possible to fabricate the larger tanks on the mainland. Standardized parts would facilitate assembly and checkout.

The delivery rate for launch stages cannot be specified in advance. It will be necessary to acquire a fleet capable of making over 200 launches per year of various kinds, irregularly distributed in time. In addition, it will be necessary to replace accidental losses and retirements. Neither of these is predictable in advance, although it seems reasonable to expect that after the early years each launch rocket will on the average survive at least 100 launches. A capacity of four units per year, operated at two units per year, may be about the right size.

The check list of missions, Table 4.2, gives a working estimate of the number of nose cone assemblies expended per year. The production rate is summarized below, in Table 6.3.4. The number of pumps, each supporting 200,000 lbs of thrust, for the launch and upper stages, is 766 per year. The number of pumps for the 50,000 lbs thrust stages is 154.

The cost of the proposed program will be mainly one of human
It will include some material resources, such as factory


[blocks in formation]

facilities, office buildings, land, a few thousand tons of iron and aluminum, a fraction of one percent of the production of methane, and a few other odds and ends, but by far the greater part of the cost will be salaries and wages. The cost can only be measured in dollars. In actual fact it is labor. The distinction is important, for it enables us to schedule expenditures in accordance with the well known characteristics of population growth curves.

Bar graphs are sometimes used to schedule the beginnings and endings of a set of interdependent tasks. Actually, each bar should be replaced by a population growth curve, showing the incubation, youth, maturity, decline, and death of the particular payroll represented. When the superposed growth curves are added together, they should give a combined population growth curve which is free from peaks and dips. If peaks and dips are present, they can be eliminated by rescheduling. A schedule thus rationalized beforehand is prerecuisite for efficient utilization of employees, and for predicting costs. It is an effective management tool, since

[blocks in formation]

the slightest departure up or down from the schedule can be immediately detected abd corrected.

Figure 6.3.2 shows in a generalized way the cost of a large, complex undertaking plotted against time to completion. Point 1 corresponds to minimum cost, and Point 2 to minimum time. The problem, of course, is to plot the curve accurately, especially before the undertaking is started. Point 2 corresponds to a "crash" program, which is a fair, descriptive title for the APOLLO moon program. In order to meet the schedules, back-up programs must be funded, organizations must be over-staffed, and excessive expediting is required. Small errors in predicting the plotted curve result in large over-runs in cost, and at least some over-run in time. Point 1, on the other hand, is much less sensitive to errors in prediction. Some urgent requirement, such as a military emergency, may dictate selection of Point 2. A solar-exploration program is best planned at Point 1. A subjective estimate is that the selection of Point 1 for a future program will reduce costs at least twofold compared with the selection of Point 2.

An exact cost prediction for the proposed program is yet to be completed. However, there are several important factors which, will make the cost less than people have learned to expect. The first of these is job experience. Assuming that the nation continues space exploration, the most important benefit of space expenditures to date will have been, not the landing of then on the moon, or the sending of spacecraft into interplanetary space, rewarding as these feats may be, but the creation of a national reservoir of skills and knowledge which can be applied to a future program. Another factor is the proposed recovery of all launch rockets, and the standardization of launch rockets, space rockets, spacecraft, and operations. A third factor is the proposed use of the new facilities, rockets, and organization for many missions over a period of many years. Ultimately, of course, we can expect to see the proposed plan become obsolete, but the usual time constants for innovation in space rocketry are measured in decades, owing to the length of time it takes to replace established teams with young people not committed to the old ways of doing things. A fourth factor is

[blocks in formation]




the ability now, for the first time, to start from the beginning with a previously planned program. These factors together are expected to reduce the annual cost of the proposed program to less than the present cost of NASA, and to increase the accomplishments by one to two orders of magnitude.

Cost data taken from the accounting records of past space programs reflect none of the factors listed in the above paragraph. In order to initiate the reiterative planning process, it is therefore necessary to make preliminary cost estimates based on experience in the repetitive production and operation of standardized machines. A summary estimate is given in Table 6.3.5. It is an order of magnitude lower than would at first be suggested by the accounting records of past NASA programs. It has not been proven right, but neither has it been proven wrong. The writer's júdgement is that if we have learned nothing in past programs the estimate is wrong.

Figure 6.3.3 places the cost estimate in perspective. The hypothetical benefits curve is geared to human competence expected to be generated by the space academy, Table 6.3.2. Time is necessary to generate human competence. The sooner the academy is started, the more the benefits curve will be shifted to the left in the figure.

[ocr errors]



The proposed plan is an engineering plan, and not a technology` R & D plan. It is based on what we know how to do, as defined in Section 1.3. The schedule calls for planning first, and action second, and not vice versa. For these reasons, it is expected to minimize the need for improvisation. Management is therefore expected to have time to cope with unpredictable problems as they arise, and to watch for unexpected opportunities.

« AnteriorContinuar »