Engineering topics

Rogers, Patrick

Publications and source records attributed to Rogers, Patrick.

3 records

NASA NTRS2000

Heat Sink Welding for Preventing Hot Cracking in Alloy 2195 Intersection Welds: A Feasibility Study

Two concepts, stationary cooling and trailing cooling, were proposed to prevent weld intersection cracking. Finite element analysis was used to demonstrate the potential effectiveness of those two concepts. Both stationary and trailing heat sink setups were proposed for preventing intersection cracking. The cooling media could be liquid nitrogen, or pressured air knife. Welding experiments on the small test panel with the localized heat sink confirmed the feasibility of using such a stationary cooling technique. The required cooling was achieved in this test panel. Systematic welding experiments should be conducted in the future to validate and refine the heat sink technique for preventing intersection cracking.

Yang, Yu-Ping↗

NASA NTRS1986

Distributing program entities in Ada

In any discussion of distributing programs and entities of programs written in a high order language (HOL), certain issues need to be included because they are generally independent of the particular language involved and have a direct impact on the feasibility of distribution. Of special interest is the distribution of Ada program entities, but many of the issues involved are not specific to Ada and would require resolution whether written in PASCAL, PL/1, Concurrent PASCAL, HAL/S, or any language which provides similar functionality. The following sections will enumerate some of these issues, and will show in what ways they relate to Ada. Also, some (but by no means all) of the issues involved in the distribution of Ada programs and program entities will be discussed.

Rogers, Patrick↗

NASA NTRS1986

Implementing distributed Ada for real-time applications

The discussion of applying a (distributed) High Order Language (HOL) to applications which require real-time performance invariably invokes the subject of excessive overhead. In a related paper, some of the basic language-specific issues involved in distributing a High Order Language, with special attention paid to the Ada language, are discussed. In a traditional implementation, several of these issues imply considerable, if not prohibitive overhead. An implementation strategy is introduced which promises to deal with these issues in a manner that will provide significant performance improvements. These improvements should in fact be sufficient to make use of distributed Ada feasible even in highly-constrained application domains. Additionally, the general approach should be applicable to nondistributed implementations as well.

Rogers, Patrick↗