Interpretation ID: nht76-5.47

DATE: 11/10/76

FROM: AUTHOR UNAVAILABLE; R. L. Carter; NHTSA

TO: Department of California Highway Patrol

TITLE: FMVSS INTERPRETATION

TEXT: Thank you for your August 12 and September 28, 1976, letters to our Office of Standards Enforcement concerning possible noncompliance of certain air-braked school buses with Standard No. 121, Air Brake Systems. At issue is the functioning of a "double check valve" between the two tanks of an air brake system designed to meet S5.7.2 of the standard (as it was in effect prior to September 1, 1976). The valve serves to provide air pressure from either supply reservior to hold off the spring brakes while guarding the air brake system from air loss through either one of the reservoirs. The identified problem stems from construction or installation of the valve such that it does not necessarily operate to seal off the brake system from a loss of air in one of the tanks, permitting loss of air from both reservoirs unless the rate of air loss is substantial. Your letter advises that the State of California is permitting the continued operation of the school buses in question pending receipt of a response from the National Highway Traffic Safety Administration (NHTSA).

I would like to address the findings you raise in the order in which they are listed on page two of your letter. I interpret your first finding to be that the "split system" designs used in compliance with Standard No. 121 may be incapable of meeting the stopping distance requirements set forth in S5.7.2. You question whether S5.7.2 adequately specifies a "back up" braking capability that is not cancelled out by a single failure of air pressure components, citing cases where a substantial rate of loss from one tank (introduced as a failure in accordance with S5.7.2.3) draws down the air pressure in both halves of a split system.

The general requirement of S5.7.2.3 (of the now superseded text) is that the vehicle, in one out of six attempts under specified conditions, must be capable of stopping from 60 mph within a distance of 613 feet "with a single failure in the service brake system of a part designed to contain compressed air or brake fluid. . . ." To provide this performance (and meet the other requirements of S5.7.2) manufacturers have provided "split systems" modulated by the service brake control. As a matter of general compliance with this requirement, you have asked if any single failure (other than failure of common components) can be introduced into the system as a test of its compliance. The answer to your question is yes. The NHTSA does not know the basis of Ford Motor Company's contention that the "failure" introduced in the system must be at least 50 psi/min from an initial reservoir pressure of 120 psi.

A noncompliance with S5.7.2.3 occurs only if the vehicle, with the failure introduced, is incapable of stopping within the prescribed distance in one out of six attempts. Thus, an extremely small failure which is easily overcome by compressor pressure would be one way of testing for compliance with S5.7.2.3, but it would not, in all likelihood, result in the demonstration of a noncompliance. As you noted in the second paragraph of page 3 of your letter, a noncompliance with stopping distance requirements may depend (largely because of a compressor's ability to overcome air loss) on the rate of leakage introduced and the time allowed between introduction of the failure and application of the emergency brake. The standard does not (except for the provision of S6.1.14 for towing vehicles manufactured on or after September 1, 1976) provide detailed specification of procedures, particularly the time between failure and brake application. The agency therefore must utilize a reasonable procedure that does not unfairly test the system. In this case, the agency considers the introduction of any size leak, followed by brake application as soon as the low pressure warning activates, to be a reasonable procedure, approximating what would occur in the event of actual failure on the highway. The agency would find a noncompliance with S5.7.2.3 if the school buses are incapable of stopping within the required distance when the described procedure is followed.

Your second finding is that a particular double check valve installation can cause noncompliance with S5.7.2.3. Consistent with our general discussion of compliance with S5.7.2.3, stopping tests are the means to discover whether the buses in question comply. We are analyzing your data, and a noncompliance investigation has been opened. We have notified Ford Motor Company of its commencement. We have also forwarded data to our Office of Defects Investigation to see if the check valve problem constitutes a safety-related defect apart from the issue of compliance.

Your last three findings address the separate question of whether the concept of a split system is adequately interfaced with the parking brake requirements of Standard No. 121. First you make the general point that, if S5.7.2.3 permits a substantial loss of air from both sides of a split system, the benefit of redundant lines is negated and the split system requirement is unjustified. The agency understands your position and believes that a stricter control on the amount of permissible leakage (perhaps by means of a limitation on compressor replenishment or a longer period before emergency brake application) may be justified. Particularly important is your point that a compressor loses capability as it grows older, and that this is not accounted for in new-vehicle tests. At the same time, however, larger failures do occur and we continue to view the split system as extremely important for these cases. The split system provides extremely good protection against failures such as rupture of a line or brake chamber diaphragm.

With regard to your point that split systems do not guard against failure of components common to the two systems as they are presently constructed (such as the compressor drive belt), I must agree with your point that no brake system can guard against every conceivable failure completely. It is for these cases, in fact, that the agency considers the automatic application of the parking brake system beneficial following loss of the other two capabilities.

You make the further point that, in cases of marginal compliance (i.e., the compressor can replace most but not all air loss from both sides of the split system), the gradual loss of system pressure permits parking brake application that interferes with modulation of the emergency capability. While S5.7.2 was designated as "Modulated emergency braking system", no specification for modulation was set forth in its requirements. The agency intended that a system conforming to S5.7.2 would be controlled by the driver instead of by automatic application. Our review of the systems you tested indicates that they are "modulated systems" as contemplated by the agency.

The present emergency brake requirements only state that the system "be applied and released, and be capable of modulation, by means of the service brake control." While further specification of this requirement may be in order, I believe it is the early application of the parking brake that actually concerns you, as it affects the application and release of the emergency braking capability.

You make the observation that early application of the spring brakes in response to air pressure loss permits them to drag, become overheated, and fade, making them useless before they can be utilized. A related issue is that the spring brakes will apply shortly after emergency brake availability (even before activation of the low air pressure warning), immobilizing the vehicle with no capability to release the parking brakes. You suggest implementation of a parking brake arrangement that would keep the spring brakes off longer or provide an isolated source of air pressure to permit their release when they do lock up.

Analysis of your test reports leads me to agree that some specification to limit the early application of spring brakes would be in order. Of course, as you are well aware, notice to interested persons of any change in the standard is required, along with an opportunity for comment. I believe that such rulemaking could be joined with the earlier rulemaking undertaken in response to a California Highway Patrol (CHP) petition (Docket 75-16, Notice 04).

Your final point is that, as long as early application of the parking brake is permissible, an isolated tank of air should be available to permit release of the parking brake from the driver's position. With regard to an isolated tank, the NHTSA continues to maintain its view set forth in our November 29, 1974, letter to Donald Gibson of the CHP that the second side of the split system provides more capability than the old systems (with an isolated tank) to avoid a lockup following a service brake failure. However, you have clearly demonstrated that the capability can be essentially negated by early application of the parking brake. I believe that limitations on early application of the parking brake would be a superior correction to the problem than the addition of more components (and complexity) to the existing systems.

Because school buses are involved in the problems you cite, you are no doubt aware that @ 103(d) of the National Traffic and Motor Vehicle Safety Act, while preemptive of State regulations or laws of general applicability that are not identical to Federal standards on the same aspect of performance, does not prevent a State or political subdivision from specifying a higher level of brake performance in vehicles "procured for its own use." Thus, the State of California may wish to order school buses with the additional isolated reservoir that you have recommended. The addition of a third tank to a system that complies with the standard's requirements would not be prevented by @ 103(d).

I appreciate the constructive approach being pursued by the State of California in enforcement of Standard No. 121. Our Office of Standards Enforcement will keep you advised of the results of its investigation.

Sincerely,

ATTACH.

DEPARTMENT OF CALIFORNIA HIGHWAY PATROL

September 28, 1976 File No.: 1.A218.A3107

Francis Armstrong -- Office of Standards Enforcement, National Highway Traffic Safety Administration

Dear Mr. Armstrong:

In this Department's letter of August 12, 1976, we brought to your attention a problem with the air brakes on school buses built on Ford chassis. We since have allowed the buses to temporarily continue in operation pending a decision by your agency as to their compliance with FMVSS No. 121.

At a meeting on September 2, in Sacramento with engineers from Ford Motor Company and Bendix-Westinghouse, this Department took the position that the Ford system did not operate as required by FMVSS No. 121 and did not protect the brakes against a single failure. Ford contended that the system was safe and complied with FMVSS No. 121. They further were of the opinion that a leak in a hose, fitting, or diaphragm of less than 50 psi/min from an initial reservoir pressure of 120 psi was not a "failure" as contemplated in FMVSS No. 121 and, in any event, could be overcome by the continuous operation of the air compressor.

The matter was pursued in considerable detail with no meeting of minds. It was found that the vehicle manufacturers' conclusions were based solely on laboratory and static tests and that no failures involving a continuous leak had been simulated on a moving vehicle. They were of the opinion that the usual vibration of the moving vehicle might result in the check valve responding more rapidly in normal service than in static tests. Although we believed the buses did not comply with FMVSS No. 121, we did not desire to keep them out of service if it could be reasonably avoided. On the other hand, we did not wish to allow them to operate if they were imminently unsafe, even if they complied with FMVSS No. 121.

We immediately scheduled dynamic tests for the following day to obtain some of the answers. Our investigation showed that the system did not function any better in tests of the moving vehicle than in stationary tests. The same situation was found to exist on other makes of school buses using similar designs. In our opinion, the systems on these buses do not comply with FMVSS No. 121 or are otherwise deficient for the following reasons:

1. It is highly doubtful that the vehicles can meet the emergency stopping distance in FMVSS No. 121 with a single failure in the air system.

2. Some double check valves are so mounted that they shuttle rapidly in one direction and are essentially inoperative in the other direction.

3. The emergency system is not capable of modulation by the service brake control as required by FMVSS No. 121 when loss of air in the service brake reservoirs results in the spring brakes beginning to apply.

4. If the driver does not react rapidly to a low pressure indication, the spring brakes can drag and overheat so as to be ineffective when they are eventually fully applied.

5. Upon failure of a compressor belt, as in the Martinez accident, the depletion of air in the service brake system upon a few brake applications could cause the spring brakes to lock with the vehicle in a hazardous location such as in a tunnel or on railroad tracks.

The results of our tests on Ford, International and Gillig buses are shown in the enclosed report. We found that all of the buses had insensitive double check valves between the primary and secondary split system. Although the buses were released for operation by the school district, we still have serious concern that this type of system does little good as an emergency brake system.

We ask that you give immediate consideration to a recall investigation to determine if air brakes using the double check valve piping system comply with FMVSS No. 121 and have an effectively-operating emergency stopping system. Our request is based on the following observations during our dynamic tests:

1. The double check valves are not sensitive enough to protect one-half of the dual system against either a mild or substantial air leak in the other half. A leak through a broken hose at a brake chamber with either partial or full foot-brake application did not cause the check valve to shuttle. If the vehicle was going downgrade with the brakes applied, the air would leak out of both systems down to the pressure at which the compressor could keep up with the leak. This ineffective dual system is a waste to the bus or truck operator, as he is no better off than with a single system with no double check valve.

2. It appears questionable whether the vehicles will meet the 60 mph, 613-foot stopping test of Section 5.7.1 of FMVSS No. 121 with a leak equal to the maximum before shuttle movement when the reservoir pressure is stabilized at the compressor output volume. If the vehicles in fact meet this performance requirement under these conditions, the "dual" components are a nonfunctional appendage. With this size leak, the pressure in the "protected" system will eventually be no higher than that in the failed system. Passage of the stopping distance test apparently depends entirely on how quickly the brake pedal is applied after the failure is initiated.

3. Double check valves are installed at the factory in mounting positions that interfere with their operation and any effectiveness that they might have. Some valves were installed with the shuttle axis vertical so that a leak needed to shut off air flow to one reservoir is far less than that at which air flow is shut off to the other reservoir. One ball-check valve was mounted at about a 45-degree angle so the ball immediately shut off a very small leak on the down-hill side but would not shut off the maximum leak that we could reproduce on the uphill side. As a vehicle gets older, contaminates and corrosion products in the system could cause the normally inactive double check valve to hang up when most needed. Even in our tests on a comparatively new vehicle, one of the check valves did not operate nearly as well upon the first simulated leak as on subsequent ones.

4. When a leak occurs, both systems drain down to the point where the spring brakes begin to drag just before the low air pressure indicator comes on. If the driver does not react to the signal, the spring brakes drag more and more as the pressure slowly drops, thereby overheating the linings and drums and destroying the effectiveness of the brakes when they finally are completely on.

5. The driver cannot modulate the brakes with a moderate leak in a supply line in either side of the dual system. The air pressure drops to the point where the spring brakes come on, even with no brake application by the driver. He can then apply the brakes harder by the foot valve, but he cannot release them below the force applied by the spring brakes. The FMVSS No. 121 amendments that became effective September 1, 1976, removed all reference to automatic emergency stopping systems which, if used, were previously required to be releasable by the driver after at least one full application. This feature apparently was intended to be taken care of in the current standard by the modulation requirement. In the systems we tested, the spring brakes could automatically apply with no action by the driver and then could not be released by the modulated treadle valve.

We realize that no system can protect against every conceivable possibility and that compromises are necessary because economical protection against one problem may result in an unavoidable creation of a smaller problem. A brake standard cannot meet every need of users and manufacturers under every possible situation, but the standard should be precise enough so users can follow its requirements and are aware of the limited assumptions on which it is based. To this end, we offer the following items that came to light in our tests. If during your investigation, you also find them to be a problem, you might wish to propose solutions to the Office of Crash Avoidance for consideration during the next revision of FMVSS No. 121.

1. Ford Motor Company contended that a leak through a damaged line, hose, diaphragm, etc., was not a "failure" as intended in FMVSS No. 121 unless the line or hose completely severed to produce a pressure drop of at least 50 psi/min from an initial reservoir pressure of 110 psi. We believe a much smaller initial leak of, say, 6 to 10 psi/min should be considered a failure for which the emergency system must compensate. With these differences of opinion, it is essential that the size of leak that constitutes a failure be specifically defined by NHTSA.

2. The stopping distance test conditions of the emergency system should be spelled out in more detail in FMVSS No. 121. The systems we checked would not isolate the protected half of the dual system under the more common failures of hoses and diaphragms. Also, the functioning of the double check valves was dependent upon their mounting position. A manufacturer's test that might be conducted to show compliance with the stopping distance standard with one system completely vented and the other with full air pressure (as would be assumed if the check valve was sensitive) is not realistic.

3. Ford Motor Company contends that their dual system is modulated as required by the standard. We contend that the system is not modulated, or only partially so, when an air leak developes which stabilizes the pressure at less than the 68-psi spring brake initial pressure. The regulation should spell out test conditions sufficient to control the intended degree of modulation.

4. The Ford system depends on compressor output to limit the extent of pressure drop in the system and, consequently, the amount of uncontrolled spring brake application. A new compressor with a 12 cfm rating can, when in good condition, keep up with a considerable air leak, but the emergency stopping system should not be dependent upon this factor. Some compressors are almost too small for some systems and many lose much of their capacity after they are in service over a period of time.

The deficiencies in the present split-air system would not be so important if a protected air tank was added to allow the spring brakes to release after the automatic application that occurs when one leak causes the air to be drained from both systems. The driver would then be able to release the spring brakes to remove the vehicle from a potentially hazardous location. Other modifications might be desirable such as the use of pressure protection valves shown in the enclosed Bendix-Westinghouse school bus system which was developed in accordance with California law a few years back.

I am sure you are aware of the intense national interest in school bus defects since the Martinez charter bus accident in which 29 persons were killed. The news media is sensitive to matters of school bus operation and may request information concerning our evaluation and recommendations to your agency. The possibility of such inquiry provides additional reason for your expeditious review and resolution of this matter.

I urge you to take immediate steps to require air brake systems of the types described in the enclosed report to be redesigned to comply with FMVSS No. 121 and to effectively protect the vehicles against brake failure.

Very truly yours,

G. B. CRAIG -- Commissioner

Enclosures

DEPARTMENT OF THE CALIFORNIA HIGHWAY PATROL

INSPECTION OF DUAL AIR BRAKE SYSTEMS ON SCHOOL BUSES

ENFORCEMENT SERVICES DIVISION

SEPTEMBER 1976

TABLE OF CONTENTS Page ABSTRACT 1 BACKGROUND 1 PROCEDURE 4 TEST RESULTS 4 I. Tests of 1976 Bluebird School Bus on a Ford Chassis 4 II. Tests of 1976 Superior School Bus on a Ford Chassis 6 III. Tests of 1975 Ward School Bus on an International Chassis 8 IV. Tests of 1976 Gillig School Bus 10 DISCUSSION 11 CONCLUSION 14

[REPORT OMITTED]