Understanding casting factors in aircraft components.
A major obstacle preventing the selection of premium quality castings for commercial aircraft structures is the subject of casting factors (CFs). Airframers hesitate to use castings because of assumed inconsistent mechanical properties and quality. Therefore, the CF was defined.
The CF is a number usually ranging from 1.0-2.0 that is attached to a casting based on its criticality. For instance, a noncritical casting might have a CF of 1.0, which doesn't present problems for the airframer in terms of design. But a critical casting with a CF of 1.5 presents large design problems for an airframer who needs to provide certain mechanical properties while minimizing weight.
In equation form, design property strength = the material allowable/casting factor. In other words, if a casting has a CF of 1.5 and the allowable for the material is 45 ksi yield strength, the maximum design property would be only 30 ksi.
The CF's continued presence in the FAR 25 (Federal Air Regulation) is sufficient to cause engineers--both experienced and newcomers--to believe all castings are unreliable.
In his 1994 Hoyt Memorial Lecture (see modern casting, July 19941, John Jorstad, CMI Tech Center, Inc., said that until the early 1980s, U.S. aircraft manufacturers accounted for 95% of the world market for commercial jets. That U.S. position has diminished to 65-75% today.
Domestic airframes now use forgings, while European airframes consist of 30--40% castings, Jorstad said.
If the CF were eliminated and castings were used extensively, structural costs of producing aircraft could be significantly reduced. For example, if entire fuselage sections were to be cast in one piece, production costs could be cut by about 30%. Considering fuselages represent up to 40% of the overall structural costs, on a $200 million aircraft, the substantial savings is obvious.
Better Casting Control Today
Looking back, the perception is that CFs were established for use with castings with only "standard" quality and consistency. When the CF was developed, statistical process control (SPC) concepts hadn't been developed.
Today's premium quality castings are manufactured and inspected using much more refined and tightly controlled processes. The ability to nondestructively examine and relate microstructures of aluminum alloy castings to strength and ductility potentials is a significant improvement. These techniques, along with the classical inspections for SPC manufactured parts, provide more product reliability than thought possible when CFs were first required.
Upon examination of FAR 25, the requirement for CFs is based on potentially appreciable strength variabilities due to uncertainties in manufacturing processes or inspection methods. That requirement is not specific. It doesn't state which strength: tension, compression, shear, bearing, fatigue, etc. As a result, many believe that CFs should be applied to all strengths. This is just one of the problems associated with CFs. Another problem is encountered when trying to find the correct factor.
For example, when airframers decide to use a casting, they must meet federal requirements regarding inspections, a major cost in the price of castings. To hold down costs, airframers may be willing to use a higher CF as a tradeoff for not making as many inspections. Unknowingly, they are violating federal requirements, since casting class and inspection requirements (per MIL-STD-2175) have not been satisfied.
Actually, airframers must satisfy all of the requirements contained in four standards when using castings:
* MIL-STD-2175 class, grade and inspection for two categories of application;
* FAR 25 Casting Factors;
* MIL-HDBK-5 design property specification controls;
* Casting procurement specification for mechanical, physical and chemical properties.
How these four standards relate to each other is the main point of confusion. For both casting producers and their customers, a clear understanding of regulations is essential to limit manufacturing and inspection (both nondestructive and destructive) costs to those that are absolutely necessary. Use of the correct CF will ensure that structural weight is minimized--a vital concern among airframers.
Confusion among Regulations
Table 1 illustrates the four regulations and how they don't relate to one another. Many engineers recognize only two of these: FAR 25 and MIL-HDBK-5. FAR 25 contains CFs that must be applied to design properties contained in MIL-HDBK-5. A deeper look at each standard TABULAR DATA OMITTED and what it controls can help clear up this problem.
MIL-STD-2175--This regulation identifies four strength classes: 1, 2, 3 and 4. Everything hinges on these four classes. Class 1 represents the most critical applications. Failure of the casting would endanger the lives of operating personnel (no mention of passengers) or cause loss of the aircraft. Failure of a class 2 casting would result in significant operational difficulties, but not catastrophic. A class 2 failure would require downtime maintenance to remedy the problem.
A class 3 casting failure would be repaired or replaced during a scheduled downtime maintenance. Failure of a class 4 casting, which aren't strength designed and wouldn't impede safe flight or landing, are least critical. With scheduled periodic maintenance, such a failure would not require fixing until the next scheduled maintenance. Classes 3 and 4 are described by margins of safety (MS)--the military term equivalent to CFs. For example, an MS of 200% is the same as a CF of 2.0.
Casting qualities are defined by four grades--A, B, C and D--which relate to classes through levels of application. It is important to recognize the relationship between grade and application categories. Minimum acceptable qualities for class 1 castings are grade B in critical areas and grade C in noncritical areas.
Grade D is acceptable for castings in classes 2, 3 and 4. The application category uses grades A-D to define criticality of intended usage. The term "critical" defines both important, highly stressed zones of castings as well as parts of a structure.
Finally, MIL-STD-2175 defines inspection requirements for each class. Other than the use of MS to define class 3 and 4 castings, this regulation doesn't mention CFs nor does it contain information on mechanical properties.
FAR 25--This regulation addresses two subjects. One is application in which the two categories of critical and non-critical are defined in much the same words used for classes in MIL-STD-2175. Neither classes nor grades, however, are cited in this regulation. CFs are presented, appearing to allow selection of factors to suit one's needs to minimize inspection costs, which is an erroneous assumption. There are also no provisions for using critical castings with CFs greater than 1.5 or less than 1.25.
MIL-HDBK-5--This regulation contains only design mechanical properties, and is totally disjointed from both of the above regulations. It lists properties by classes, but they are not the same as those identified in MIL-STD-2175. Such classes correspond to those listed in procurement specifications describing different strength levels.
Allowables do not state applicable grades, causing confusion among engineers. It is also beyond the scope of MIL-HDBK-5 to suggest what type of application is appropriate for any material. Finally, no mention is made of CFs. All properties contained in the handbook are pristine, without CFs.
Specifications--Specifications classify guaranteed minimum tensile properties by classes that refer to strength, not application criticality. ASTM E155 grades are defined in most specifications. But it isn't clear in all cases which grades relate to guaranteed minimum properties. Use of the terms "critical" and "noncritical" refers to zones of castings, not types of application. These casting zones might also be described using the terms "designated" and "non-designated" or "other" areas.
An orderly process is needed to enable selection of the applicable class, grade, types and levels of inspection, and appropriate CF to suit each intended application.
Selecting the Casting Factor
Figure 1 outlines the process that should be used to select CFs. The process does not start with FAR 25--it begins with the intended application. Based on use, one must immediately refer to MIL-STD-2175. The intended application dictates which of the four strength classes apply in using a casting. In turn, class sets the grade, types and levels of inspection and CFs that must be used.
Only class 1 applications require that the castings be of high quality. While critical areas of such castings must be grade B or better, noncritical areas of the same castings may be grade C or better. Grade D castings may be used in all other applications.
In other words, if the failure of a casting won't endanger safe flight or lives, it would be inappropriate and unnecessarily costly to specify grade A, B or C castings. In today's production environment, however, it may be difficult for many foundrymen to produce less than grade C castings without abandoning their controlled manufacturing processes.
Class also sets types and levels of inspections. Inspections can't be selected to satisfy cost. For class 1 castings, all parts of all castings must be thoroughly examined by visual, magnetic particle (ferrous casting) or penetrate (nonferrous castings) and radiographic (or suitable equivalents to film x-ray) inspections. Classes 2 and 3 require fewer inspections, based on acceptable sampling plans. Class 4 castings, which are not intended for structural use, do not require radiographic inspections.
Once inspection types and levels have been established, this leads directly to CFs contained in FAR 25. Exactly the same information as shown in Table 1, inspection establishes which CF should be used. This is not a multiple-choice opportunity. Once application is defined, all other issues fall into place.
There appears to be a loophole. No mention was made for the specific alloy or its design mechanical properties. Selection of the proper CF, however, doesn't require prior knowledge of the specific alloy or its properties.
At this point, selecting the CF boils down to MIL-HDBK-5, the source for allowable design properties. With one exception, statistically based casting allowables do not exist. For all alloys (other than D357.0-T6 Designated Area castings), MIL-HDBK-5 contains only guaranteed specification minimum requirements (S-values). S-values have no specified levels of confidence or probability. They represent what foundries are willing to guarantee in tested castings.
It is no wonder CFs are still required. Without statistical assurances for properties, it is difficult to argue that uncertainties in strength have been eliminated. For the foundry industry to grow in the airframe market, it must become more active in MIL-HDBK-5 meetings and get statistical A- and B- basis values developed for their products.
Over the past few years, however, only a handful of foundries has sent representatives to these meetings. Foundries must accept the burden for proving that uncertainties ascribed to manufacturing processes and inspection methods have been eliminated.
The second item of concern is the lack of allowables contained in MIL-HDBK-5 for titanium casting alloys. This is not meant to imply that such properties do not exist, since they are specified in certain company design manuals. Not all companies, however, feel they can afford to produce their own design manuals.
Eliminating Casting Factors
When a number of requirements is satisfied, a unity CF is already permitted for noncritical castings. Parts must be procured from a specification that guarantees properties. An approved sampling plan must be used to periodically cut up parts to prove guaranteed properties are met. All parts must be thoroughly inspected by nondestructive methods. Three castings must be statically tested to show they meet certain strength requirements under simulated service loadings.
The Air Force has successfully used D357.0-T6 aluminum alloy castings in a few applications without CFs. This was a well-executed plan, based on information generated under AF Contract F33615-85-C-5015 and reported by Ozelton, and others, in AF Report WL-TR-91-4111, (October 1991). Castings were provided by three premium quality foundries.
To describe how conservative the statistically based A- and B- values are for designated areas of this alloy, this report contained results of 183 tensile tests, describing 21 conditions. The data base includes the following conditions:
* grades A, B, C and D;
* designated and non-designated areas;
* sodium and strontium modifications;
* welded (grades A/B) and unwelded areas;
* gas porosity producing grades B, C and D;
* shrinkage producing grades B and C;
* foreign material inclusions producing grades B and C;
* unacceptable microstructures.
With the A-value (lower limit allowable) at 39 ksi yield strength and 46 ksi ultimate strength, the average of all the above conditions showed an average 44.2 ksi yield strength and 51.8 ksi ultimate strength. Specification guaranteed minimums (S-values) are 40 ksi for yield strength and 50 ksi for ultimate strength. However, it was decided these allowables would be applicable only to castings that were: unwelded, in designated areas and had inspected microstructures that were determined acceptable.
In other words, by taking every deleterious variable imaginable, it was found that average strengths were still well above minimum requirements. Then, computed A-values were limited to only certain favorable casting conditions. With this amount of control and conservatism, there should be no need for CFs.
Figure 2 shows a normal distribution plot of minimums reported for each of the 21 conditions. Only one of the ultimate strength results (inclusion, grade C) falls fractionally below the statistical A-basis value of 46 ksi. This is insufficient to warrant the need for a CF.
MIL-STD-2175 describes classes 1 and 2 with reference to "failure" and the resulting effect. FAR 25 describes critical castings in terms of "failure." The commercial air transport view of failure is that ultimate strength was exceeded.
Many people consider failure of the part to mean it is no longer there--only a gaping hole remains. If this is part of the pressurized envelope of an airplane, then criticality is determined by the part's square footage. Foundrymen need to understand how the airframer considers the application of each casting.
Tension or flexural failures may only represent elongated cracks, similarly to those experienced in other types of metallic structures. Such parts are repairable during maintenance using existing methods of welding or plugging with mechanically attached and adhesively sealed load-bypass plates. "Failed" castings simply don't vanish. Compression or shear instability failures would lead to excessive deformations, not disappearance of the casting.
The subject of CFs is more than just a domestic issue. Foreign regulations published by the Joint Aviation Administration in JAR 25 contain the same information as found in FAR 25. Therefore, FAR 25 will not be changed without approval of both the FAA and JAA, whose goal is to establish and maintain harmony.
Foreign airframers have been working with their regulatory agencies for years to alleviate CFs. So, a clear understanding of current requirements is essential before any domestic plan for alleviation can be established. If the major use of castings continues to be for noncritical applications, FAR 25 already allows the use of a unity factor upon satisfying specification, coupon properties, inspection and full-scale test requirements. In this case, alleviation is not required.
|Printer friendly Cite/link Email Feedback|
|Article Type:||Cover Story|
|Date:||Oct 1, 1994|
|Previous Article:||180 foundrymen explore spent sand options.|
|Next Article:||Assistant editor trades computer for ladle and rammer.|