Exhibit VF-2. Retype of memo (with GAS notes added). The Classification of this memo and its enclosures was cancelled.

From: RAEV

To: CD-3

Subj: F-111, Review of Project Initiation

Ref:
(a) Secret Memo RAEV to R dtd 9 Jan 1963
(b) DDR&E Secret Memo for SecDef dtd 19 May 1961 (Note: Signed by Dr. Brown)
(c) BuWeps to CNO Secret R-5:FMG dtd 3 May 1961
(d) SecNav Secret Memo for SecDef Ser 006038P50 dtd 22 Aug 1961

Encl:
(1) Table - History - TFX Characteristics dtd 2-2-65
(2) Chart - Unit Cost Data - 1961
(3) Chart - Unit Cost Data - 1964
(4) Chart - Comparative Unit Costs - 1964 vs. 1961
(5) Chart - Cumulative Costs

1. The F-111 program has been a controversial one from at least two standpoints. First, the program decision was made in the face of opposition by both services since neither the Navy nor Air Force believed a single design could meet their stated requirements. Second, the final source selection decision was contrary to the unanimous military recommendations. The role of BuWeps in the selection process has been documented for the record by reference (a). The fact that the first airplane has now flown is being taken as proof that the services were wrong in their prediction of a lack of technical feasibility. Coupled with the technical feasibility, there has been the question of economic feasibility. Large cost savings for a single program were predicted. For the record, a review of the facts on which the early decisions were based is considered necessary.



2. Early in 1961, Project 34 was established by SecDef to review the overall problem of tactical type aircraft in the 1962-1971 time period. The final report on that project, reference (b), recommended a single TFX project, under Air Force administration, to meet the air superiority needs of the Navy, the tactical requirements of the USAF and for CONUS defense. Also recommended was the VAX to be developed by the Navy for close air support after further studies. Prior to issuance of the report, working groups had developed service positions while additional data were generated by a NASA-WSEG-DOD group. The designs of the latter group were considered unreasonably optimistic by the BuWeps Working Group as reported in reference (c). (Note: BuWeps "Working Group" - most work done by Gloeckler and his division, I was not yet involved. GS 1990)
   
   The Project 34 report shows that a single compromise design would be one billion dollars less expensive than the individual programs recommended by the Air Force and Navy. Although more studies were conducted prior to the final SecDef decision about three months later, Project 34 appears to be the real starting point of the F-111, and the source of the "Billion Dollar Saving." The basic characteristics on which the Project 34 recommendation was based are tabulated on enclosure (1). The first column shows the Navy recommended design, the second column the Air Force recommended airplane, while the DOD recommended compromise is shown in the third column. The characteristics and costs of the Navy and Air Force designs were provided by the respective services. The characteristics of the compromise design were apparently of DOD origin, while its costs were said to be based on Air Force TFX data. The following characteristics should be noted:
   
   
   1. All designs used two TF-30 engines.
   2. The Navy design emphasized holding size and weight to a minimum. The 56 ft. long airplane with a gross weight of 50,000 lb. carrying 6000 lb of missiles and 17500 lb. of fuel was questioned as being optimistic by DDR&E.
   3. The Navy design had a radius of 555 miles on a Lo-Lo-Hi mission with a M1.0 dash of 100 miles when carrying external fuel. This was well below the 800/1.2/200 combination required by the Air Force in their SOR-183.
   4. The basic Air Force design provided 50% more CAP time for the Navy mission than did the Navy design. However, the length of 82.5 feet made it impossible to operate on carriers. At 63000 lb. it met the 800/1.2/200 Lo-Lo-Hi requirements with no external fuel.
   5. The recommended compromise design showed an Air Force radius of only 340 miles with a Mach 1.2 dash of 100 miles. With 12000 lb. of external fuel, the dash distance was 100 miles, half that specified by the Air Force, at a total radius of 830 miles. CAP time for the Navy was more than required, while the Navy's 56 foot length was claimed to be met by folding 10 feet of the airplane.
   6. The technical characteristics of the compromise design appear fairly consistent with the other two models.
   7. Cost data for the designs appear grossly inconsistent. The heavier and larger Air Force design with a smaller buy reaches a lower unit production cost than the other two designs.
      
      
3. On enclosure(2) are plotted unit "procurement" costs excluding R&D as they are given in reference (b). It is not clear whether those costs are "Flyaway", "Program" or "Investment." The inconsistency in cost data is quite obvious. The much larger and heavier Air Force design is more expensive initially than the Navy design but reduces rapidly so that its last buy is but 60% as expensive at the same point in production. The compromise aircraft falls between the two other designs as expected but it, too, should not be less expensive than the Navy airplane. On the basis of the cost data used, it appears that a larger dollar saving would have resulted from a buy of the Air Force recommended design instead of the compromise one.



4. By memorandum dated 7 June 1961, SecDef indicated that the Air Force would be authorized to develop a new air superiority aircraft to be used by both services to replace the F4 and F105. Working committees attempted to reach agreement on the characteristics of the single design from that date to 22 Aug 1961 when SecDef was informed that it was considered not technically feasible to meet the stated requirements of the two services, and recommended separate development programs with the Navy taking advantage of the Air Force program to the maximum possible extent. The 22 Aug paper repeated the basic requirements of a 56 ft. long, 50,000 lb. airplane as those which the Navy desired. In a spirit of compromise, a 55,000 lb., 61 ft. design foldable to 56 ft. was offered as the maximum that could be accepted by the Navy. This design would have a Mach 1.2 dash speed capability, but over a 100 mile dash distance capability. The Air Force position remained firm that their full 800/1.2/200 requirement be met. The characteristics of the three approaches are tabulated in columns 4, 5, and 6 of enclosure (1). Additional study data were provided by the Navy showing that the Air Force design with its ASG-18/GAR-9 missile system would be only 37% as effective as the Navy design for fleet air superiority. The Navy offered compromise was calculated to be 78% as effective as the smaller design. Unit production prices of 3.0, 3.3, and 3.5 million were quoted in going from the "basic Navy" to the "compromise Navy" to the "Air Force" design. Details of the latter two designs are somewhat meager in reference (d), but it appears that the Air Force design must have remained essentially the same as tabulated in column 2.



5. On 30 August 1961, DDR&E recommended the single design approach to SecDef who directed implementation on 1 Sep 1961. It seems clear that DOD gave more weight to the more optimistic NASA-WSEG-DDRE studies than to the service positions, but it was recognized that a "challenge" to industry was being presented. The characteristics outlined in the decision paper are shown in the 7^th column of enclosure (1). The following points are significant:



1. SecDef directed initiation of the project within certain constraints, stating that he believed "development of a single aircraft of genuine tactical utility to both services in the projected time frames is technically feasible." Note that only tactical utility was claimed as feasible in the actual decision paper, not the meeting of stated requirements. The DDR&E paper indicated that it was feasible to meet performance requirements within the constraints imposed.
2. The Air Force version was to weigh "approximately 60,000 lbs..." while the takeoff weight of the Navy version was not to exceed 58000 (???? unreadable) 150 mile mission with 6000 lb. of missiles "without the consent of the Navy."
3. A 36-inch minimum antenna diameter was to be provided (???? unreadable) length of the Air Force version was not to exceed 75 ft. In the DDR&E paper this was to give a potential Navy length of 56 ft. by removing a section of the Air Force fuselage and folding the nose.
4. Basic performance requirements of SOR-183 were to be met as "nearly as possible" within the specific constraints imposed.
5. Specific carrier compatibility requirements were not delineated. The Navy had desired limited operation capability from the CVA-19, while DDR&E believed operation from CVA-59 and better would suffice. The controversy, documented in reference (e), resulted in apparent agreement that full operation from the CVA-43 would be required, but this was eventually changed to only limited operation from carriers below the CVA-59.
   


6. The requirements were incorporated in a "Work Statement" for industry to submit proposals. As detailed in reference (e), none of the proposals met the combined requirements in the first round submission. The designs submitted in the second round failed to meet requirements by an even greater degree, as was predicted by the Navy. After the second round, the Air Force still stated that all requirements could be met by "refining" the winning design. The Navy still considered the task impossible since correction of the deficiencies all tended toward increased weight, already too great, and increased size, already too large. The third round was then arranged to define the differences between Air Force and Navy versions which would have to be accepted if the performance requirements of the two services were to be met. The Air Force also asked the contractors to define the differences between versions necessary to reduce the weight of the Navy version to the original 55000 lb. while meeting the performance requirements. As a result of the third round submissions, it became evident that the Navy airplane would have to be greater than 55000 lb., and other compromises would also be necessary. "Acceptable" levels of compromise were defined for the fourth round for the Navy airplane, while all Air Force generated requirements stayed firm.



7. In columns (8) and (9) are shown the characteristics of the final G.D. proposal as submitted and as evaluated. It can be seen that the airplane failed to meet the Lo-Lo-Hi radius and takeoff distance on Air Force figures. From the originally Air Force recommended airplane (column 2) the weight was up 10% and even with 16% more fuel, the dash distance was down from 200 to 125. The Navy version was 27% heavier and its unfolded length 10 ft. longer than had been originally specified for the Navy. The airplane was not considered fully operational on CVA-43 by BuWeps. In comparison to the characteristics outlined in the SecDef decision, even with the gross weight constraint exceeded by 16%, the high speed dash distance in the radius problem was low by 37%. Simultaneously, the Navy version was well beyond its weight constraint, and the 56 ft. folded length potential was not achieved.



8. In the course of the development cycle, the Air Force permitted the contractor to increase airplane size and fuel tankage in an effort to meet the Lo-Lo-Hi radius requirement. It seems clear from the airplane's weight growth that both the Navy and Air Force seriously underestimated the penalties associated with the meeting of the dual requirements. It is also clear that the NASA-WSEG-DOD studies lacked any semblance of reality.
   


9. A strict comparison of costs is not possible without a major amount of research. The readily available position papers lack preciseness of terms in the cost area, and in addition, definitions have changed. It is probable that avionics costs were markedly low in the early estimates. The quantity of aircraft included in the program buys has changed with time. As noted on enclosure (1), the early 1961 estimates showed a 934 airplane program for the Navy, and 779 airplanes for the Air Force. In the 1962 proposal, 1495 production Air Force airplanes were included, but only 231 Navy airplanes. This was an artificially low number, due to the use of a five-year plan. The Navy buy is now programmed at 350 while the Air Force has dropped to 749. Care must be exercised in drawing conclusions based on average costs with such large changes in total quantities, and ratios between individual service buys.



10. On enclosure (3) are shown the current unit costs of the F-111A and F-111B with each plotted at the mid-point of the total buy. The F-111B has more expensive electronic equipment, but is procured later, relatively, that the F-111A. If each model is plotted against its own numbers, rather than the combined numbers, the unit costs of the two airplanes are quite similar. A comparison of the 1961 average curve from enclosure (2) and the 1964 average from enclosure (3) is shown on enclosure (4). It can be seen that the flattened-out costs today are at least double those on which the original Project 34 recommendation was made.



11. On enclosure (5), a cumulative cost plot is shown. The lower dashed line is the original Navy recommended airplane program. Its completion point forms the origin for the Air Force recommended program. The solid line is for the combined program as recommended (and priced) by DDR&E. The billion dollar saving is apparent. A single spot is shown for the Air Force estimate as presented in the 4^th evaluation report. It is seen that that cost is about 2 Billion above the original estimate. Coincidentally, it is about equal to the current total cost, although the number of aircraft has decreased by 35% from 1726 to 1122.



12. The changes in weight of the airplanes must be considered when examining the relative validity of the original cost estimates. A very gross comparison might be done by comparing the increase in the so-called (????unreadable) unit cost with the increase in weight of the airframe. The following table where such a comparison with the weight reference taken as gross weight (???) and ordnance.

	Navy	Air Force
1961 weight	26500	36000
1961 cost	2.8	1.5
1964 weight	45100	43950
1964 cost	4.4	3.8
Increase weight	70%	22%
Increase cost	57%	153%

It seems evident that the original cost projections of the Air Force (???unreadable) to have been used for a base for the DDR&E estimates for a single aircraft program were grossly optimistic.

13. In summary:
    
    
1. The original DDR&E recommended airplane compromised the Air Force Lo-Lo-Hi radius requirement to a marked degree. The costs of that airplane were substantially lower relative to the Navy airplane than they should have been.
2. The refusal of the Air Force to compromise its mission requirements resulted in the Navy reaching a conclusion that the single aircraft program was not technically feasible.
3. Overly optimistic design studies by groups outside the Navy and Air Force led to a technical requirement far more demanding than the original recommendation had contemplated.
4. The weight limit imposed in the 1 Sep 1961 SecDef decision was treated not as a "constraint" but as a low priority requirement.
5. Air Force insistence throughout the source selection process that all requirements could be met by "refinement" of the design undoubtedly contributed to the general confusion.



14. Decisions are based on facts. In this case it is clearly evident that inconsistent facts were presented, and time apparently prevented a reasonable analysis of those inconsistencies. This was the fault in the original decision of May and June 1961. As implementation of that decision was attempted, technical over optimism in the design state-of-the-art became dominant. Proper distinctions, due to lack of knowledge, were not drawn between expertise in the theoretical aerodynamic field and the real world of airplane design. Throughout the years, the preliminary design efforts of research organizations have been uniformly more optimistic than those done by groups linked with service experience. This caused no trouble as long as the design decisions were made within the services where full knowledge of the facts existed.



15. For the future, it is obvious that the Bureau must have the ability to make both technical and cost estimates which will be accepted at higher levels. Our estimates in the F-111, criticized because of conservatism, have actually been too optimistic.

G. A. Spangenberg

Copy to: RA-2

[ed: A similar chart to enclosure 1 that is a little simpler to read is included with this exhibit]