ON MOTION FOR REHEARING
JOHN T. BOYD, Chief Justice.
After considering appellees’ motion for rehearing, we withdraw our opinion of May 25, 1999, and substitute the following opinion.
In this appeal appellant, Turbines, Inc. presents five issues challenging a judgment against it for claims arising out of an airplane crash. In response to a jury verdict, the trial court awarded Van Dardis (Dardis), the pilot of the plane, and his wife Danna Renee Dardis, $435,455 in actual damages and $500,000 in exemplary damages for their claims of negligence, gross negligence, and strict products liability. In its first two issues, Turbines questions the submission of jury questions on strict products liability and an instruction on
res ipsa loquitur,
in issues three and four, it challenges the sufficiency of the evidence of negligence and gross negligence, and in issue five, it assigns error to the admission of evidence that Turbines had temporarily surrendered its repair station certificate. For the reasons stated herein, we reverse the judgment of the Mai court and render judgment that ap-pellees take nothing.
The nature of Turbines’ challenges require a rather detailed recitation of the evidence presented in the Mai court. Turbines is in the business of performing maintenance and repair on turbine aircraft engines. They have a repair station certificate from the Federal Aviation Administration (FAA) to perform work on engines manufactured by Pratt
&
Whitney. The certificate excluded authority to perform
certain procedures, such as overhauls. G & G AeroServ, Inc. performs mechanical work on aircraft engines and airframes. Working together, Turbines and G & G developed a procedure for modifying a model of aircraft used for crop dusting, the Piper Brave, from its original piston engine to a turbine engine. The purpose of the modification was to provide the airplane with increased power and reliability. Although the Piper Brave airframe was only rated for 375 horsepower, the PT6 turbine engine could produce 550 horsepower. Part of the modification included instrument markings to indicate the maximum power setting for the engine so the pilot would not exceed the 375 horsepower limit. G & G obtained approval from the FAA in the form of a Supplemental Type Certificate (STC) to perform this modification commercially. Dardis was an aviation mechanic certified by the FAA and a co-owner of G & G. Dardis worked primarily on turbine engines and helped obtain the STC.
Glen Carlson, a resident of Montana, owned a Piper Brave and asked G & G to convert it to a turbine engine in accordance with the STC. The model of engine selected for the conversion was a Pratt
&
Whitney PT6A-6. Turbines assembled and tested a used PT6A-6 turbine engine
in the summer of 1993 and subsequently shipped it to G
&
G. Dardis personally installed the new engine and tested it before flight. The aircraft was certified as airworthy by G & G on March 22, 1994. The following day, Dardis departed from Lubbock International Airport to return the plane to Carlson. Shortly after takeoff, Dardis noticed sparks from the exhaust and an elevated engine temperature. He returned to the Lubbock airport to investigate the problems. When these conditions were reported to Turbines, Jim Mills, President of Turbines, asked Dardis to return half of the engine, the power section,
for inspection. Turbines was unable to find any defects in the power section but replaced a bearing and returned the power section to G & G. After Dardis reinstalled the power section, he continued to express concern over an elevated engine temperature. Mills explained the temperature difference was a result of the difference in the PT6A-6 engine and the PT6A-20 model.
The Crash
On April 1, 1994, Dardis again attempted to return the modified plane to Carlson in Montana. After an uneventful flight from Lubbock, Dardis landed in Beloit, Kansas, to show the plane to agricultural pilots and potential customers. After showing the plane to these pilots, Dardis taxied onto the runway for departure. The parties present differing accounts of the events which followed.
According to Dardis, he performed a normal takeoff and the plane performed properly during takeoff. He climbed to about “a couple hundred feet” at the “best angle of climb,”
leveled out, then made a 180 degree right turn with a 30 to 45 degree bank angle. After he leveled the plane from making the turn, the airplane “just lost power.” It immediately lost airspeed and the right wing stalled
and dropped. Dardis attempted to apply full power but the engine did not respond and the plane continued to “fall.” Dardis crashed through a hangar and came to rest on the building’s concrete floor.
The eyewitness accounts presented by Turbines consisted of the testimony of three pilots who witnessed the takeoff and ensuing crash: Robert Kadel, an agricultural pilot with 8,000 hours of flight experience, 6,000 hours of which were in agricultural aircraft; Jim Rome, an agricultural pilot with 14,000 hours of flight experience; and Steve Thompson, an airline transport pilot and flight instructor -with 10,000 hours of flight experience. Each testified that when Dardis departed from the Beloit airport, he only taxied about 1,000 feet down the 3,600 foot runway before turning around to depart to the south. Each testified Dardis’s ground roll was very short and the airplane climbed at a very steep angle. They also testified that when Dar-dis made a 180 degree turn, it was a very sharp turn in which he banked the plane approximately 60 degrees. According to Rome, the bank angle further increased to almost 90 degrees, which would place the wings in an almost vertical position. Ka-del explained that a high bank angle significantly increases the airspeed required to avoid a stall. Kadel and Rome testified that the airplane stalled and rolled or went into a spin and rapidly lost altitude until it crashed into the hangar. Thompson said the plane “fell” into the hangar. Each of the pilots testified that there was no change in the engine sound during the flight.
As is apparent from these differing accounts, Dardis’s theory was that the crash was caused by a sudden engine failure, while Turbines took the position that Dar-dis did not maintain sufficient airspeed, caused the plane to stall at a low altitude, and was unable to recover control. The remaining evidence presented at trial consisted of expert testimony from post-accident investigations. We will examine the evidence of each party in turn.
Evidence on Engine Failure
In support of his theory of engine failure, Dardis presented expert testimony that the engine was not producing power at the time it struck the ground and that the cause of the engine failure was the failure of a bleed valve in the compressor section of the engine. The first expert called by Dardis was David Hall.
Hall’s expert testimony was that the engine was producing little or no power at the time it struck the ground. This opinion was based on examinations of the propeller and engine. Factors leading to this conclusion were that the propeller was bent off of the supporting shaft rather than being twisted off and that the exhaust case, which provides the structural connection between the front and rear of the engine, was “bent and kinked” but not severely twisted, as he would expect if the engine had been producing high power on impact. Hall declined to give an opinion on why the engine was not producing power at impact. On cross-examination, he conceded it was “possible for a pilot seeing the ground coming up, would close the throttle.”
Dardis also presented the testimony of Don Hamill. Hamill was a former employee of Pratt & Whitney where his duties focused on investigation of external components of Pratt & Whitney engines. The investigations were of components returned to Pratt & Whitney by mechanics because they did, or were suspected of, malfunction. Some of these investigations arose out of aircraft crashes. Hamill stated he had investigated thousands of compressor bleed valves. Turbines presented
no objection to Hamill’s qualification as an expert in the external components of the PT6-A engine.
In his investigation, Hamill ruled out other systems, such as the fuel system, as the source of any failure. When he went to examine a component called the compressor bleed valve, he found that it was missing. Because he was not able to examine the valve and had ruled out malfunction in other systems, Hamill concluded that a failure in the bleed valve caused a loss of power and resulting crash.
Hamill supported his conclusion by explaining the operation and construction of the valve, together with the known history of the valve installed on this engine. The testimony of Hamill and others explained that turbine engines consist of two sections, a compressor section and a power section. The compressor section supplies compressed air to the power section where it is mixed with fuel and burned. At low engine speeds, the compressor supplies more air than is needed by the power section and the bleed valve allows the excess pressure to be vented. The valve remains open (venting pressure) until the engine reaches about 80 percent of its maximum power, at which point it begins to close. The valve is designed to close by the time the engine reaches full power.
The valve operates by using the pressure with the differential between two points acting on a rubber diaphragm. It was Hamill’s opinion that a pinhole in the diaphragm would result in an increase in engine temperature. He further stated that the maintenance manual directed a mechanic to check the bleed valve when the engine experienced a high temperature. On cross-examination, Turbines showed that the temperatures experienced by Dardis were within normal limits and the manual did not require any action. Hamill continued to assert that a 200 degree increase in temperature should have caused Turbines to investigate the valve, even though that was not a diagnostic criteria in the engine manual. He also stated that a tear in the diaphragm would result in a loss of power of the type described by Dardis.
With regard to the particular bleed valve installed on this engine, Hamill used the serial number from the engine records to determine that the same valve was on the engine from at least 1968, and possibly as early as 1964, when the engine was built. The engine was used in an aviation mechanics school in France for several years, during which it was not used in an aircraft. It was used in one or more aircraft in the 1980s. In 1992, the engine was returned to Pratt & Whitney for inspection and repair after it was shut down in flight while being used on a King Air. After disassembly, but before inspection or repair, the engine was sold to Turbines, where it was reassembled and tested before being shipped to G & G. Hamill opined that the alternate periods of storage and use, together with age, embrittled the rubber diaphragm, making it susceptible to failure.
Through Hamill’s testimony, Turbines introduced four engine service bulletins addressing the compressor bleed valve on the PT6A engine. They were bulletin numbers 1418, 1414, 1415, and 1417, issued in 1985. Each contained a number of sections, including the reason for the bulletin, a description, whether compliance was optional, recommended or required. Each also contained specific steps for complying with the bulletin. Bulletin 1413 was listed as optional, the other three were “recommended — desireable.” Pratt & Whitney recommended performing the work “when disassembly of [the] engine is sufficient to afford access to the affected part.” According to the engine records, none of these service bulletins had been implemented. Hamill did not give an opinion that the failure to implement these bulletins fell below the standard of care for an aviation mechanic.
In controversion of Dardis’s engine failure theory, Turbines called Roy Haight,
chief inspector for Turbines. Haight described the process Turbines uses when assembling an engine. He explained that although Turbines does not perform engine overhauls, it assembles engines using the overhaul manual because it imposes stricter criteria. With regard to the compressor bleed valve, Haight stated that the fact the valve had been on the engine since 1968 was not significant because the records showed it only had 900 hours of flight time which, he stated, was a very low number of hours. After Turbines assembles an engine, it performs several tests, one of which involves running the engine with several instruments attached. During this test, one end of a hose is attached to the outlet of the bleed valve and the other end is placed in a container of water. The operator watches the bubbles produced while increasing engine power to ensure that the valve closes at the proper rate. Haight stated that the bleed valve on the engine sold to G & G operated properly in the test.
Haight also sought to explain the effect of a failure of the bleed valve. He stated that because the valve did not begin to close until the engine reached 80 percent of its total power, even if the valve failed, the engine would produce approximately 400 horsepower. The valve was designed to fail in the open position so the engine would continue to operate even if it could not produce full power. His opinion was based on personal experience in testing engines where the bleed valve had failed.
Evidence on Pilot Error
In addition to the eyewitness testimony set out above, Turbines presented the testimony of Warren Wandel in support of its position that the crash was the result of pilot error. Wandel cited the Pilot’s Operating Handbook for the modified Piper Brave to show that the stall speed in straight and level flight was 70 miles per hour. When given a set of hypothetical facts concerning the same aircraft flying at an airspeed of 70 to 80 miles per hour, as Dardis admitted in deposition testimony, then banking 30 to 45 degrees to turn, it was Wandel’s opinion that the aircraft would experience a stall.
In an attempt to rebut this testimony, Dardis presented the testimony of Dennis Way, a pilot and flight instructor. Way also cited the Pilot’s Operating Handbook to determine the stall speed in straight and level flight. However, because the aircraft was lightly loaded, he concluded that the stall speed was 63 miles per hour. Citing a chart in the handbook which does not appear in this record, Way testified that the stall speed would have increased to 77 miles per hour in a 45 degree bank and 90 miles per hour for a 60 degree bank. He suggested that Dardis’s deposition testimony as to his speed was incorrect because had he been flying between 70 and 80 miles per hour, he would have stalled at the beginning of the turn rather than at the end, where the witnesses agreed the stall occurred.
Turbines identified Dardis’s limited experience as a pilot as evidence supporting its theory of pilot error. The record showed that Dardis had a private pilot’s license with 300 hours of flight time. As a point of reference, the regulations governing eligibility for pilots’ licenses require a minimum of 40 hours of flight experience before being eligible to obtain a private pilot’s license. The flight experience requirement for a commercial pilot’s license in single or multi-engine airplanes is 250 hours. Gary Bradley stated he was against Dardis flying the plane back to the owner and that there were more experienced pilots available to return the plane. Turbines presented evidence concerning medications Dardis was taking at the time of the crash.
Turbines’ first point presents two distinct challenges, the first assigns error to jury question one asking whether the engine was defective when Turbines sold it to
G & G.
The second point challenges the sufficiency of the evidence supporting the jury’s finding that the engine was defective. In support of its first challenge, Turbines argues submission of the product defect question was error because 1) the evidence conclusively showed no defect existed, and 2) the trial court had already granted a directed verdict in favor of G & G on the same issue.
A trial court is required to submit controlling questions to the jury if the issue is properly pleaded and is supported by some evidence.
Triplex Communications, Inc. v. Riley,
900 S.W.2d 716, 718 (Tex.1995); Tex.R. Civ. P. 278. Review of a trial court’s decision to include or omit a question from the jury charge is governed by the abuse of discretion standard.
Crawford v. Hope,
898 S.W.2d 937, 940 (Tex.App.—Amarillo 1995, writ denied). That standard requires the reviewing court to determine whether the actions were arbitrary or unreasonable.
Downer v. Aquamarine Operators, Inc.,
701 S.W.2d 238, 241-42 (Tex.1985).
To determine whether there was some evidence of a product defect supporting Dardis’s strict products liability claim, we must first review the factors which may render a product defective. Our supreme court has adopted the strict products liability standard set forth in section 402A of the Restatement (Second) of Torts.
Uniroyal Goodrich Tire Co. v. Martinez,
977 S.W.2d 328 (Tex.1998);
McKisson v. Sales Affiliates, Inc.,
416 S.W.2d 787, 788-89 (Tex.1967). That section provides:
(1) One who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his property is subject to liability for physical harm thereby caused to the ultimate user or consumer, or to his property, if
(a) the seller is engaged in the business of selling such a product, and
(b) it is expected to and does reach the user or consumer without substantial change in the condition in which it is sold.
(2) The rule stated in Subsection (1) applies although
(a) the seller has exercised all possible care in the preparation and sale of his product, and
(b) the user or consumer has not bought the product from or entered into any contractual relation with the seller.
Restatement (Second) of Torts, § 402A(1965). A product may be unreasonably dangerous because of a defect in design, manufacturing, or marketing.
Uniroyal,
977 S.W.2d at 335;
Thiele v. Chick,
631 S.W.2d 526, 530 (Tex.App.—Houston [1st Dist.] 1982, writ ref'd n.r.e.).
The trial court’s question did not limit the jury to finding a single type of defect. Consequently, Turbines challenges the evidence supporting each category of product defect. It argued that 1) it did not design or manufacture the engine or its components, 2) that Dardis’s own expert, Hamill, testified there was no design, manufacturing, or marketing defect in the compressor bleed valve, and 3) there was no other evidence of a defect. Turbines’ point does not challenge the imposition of liability on it as the seller of a used product. Texas courts have long recognized such potential liability.
Mid Continent Aircraft Corp. v. Curry County Spraying Service, Inc.,
553 S.W.2d 935, 939 (Tex.Civ.App.—Amarillo 1977),
modified,
572 S.W.2d 308;
Hovenden v. Tenbush,
529
S.W.2d 302, 310 (Tex.Civ.App.—San Antonio 1975, no writ);
McLain v. Hodge,
474 S.W.2d 772, 774 (Tex.Civ.App.—Waco 1971, writ ref'd n.r.e.).
See also
Products Liability: Application of Strict Liability Doctrine to Seller of Used Product, 9 A.L.R.5th 1,1993 WL 837761 (1993).
Dardis argues that he was not required to establish what the defect was and that “evidence of a product’s malfunction will sustain a jury finding.” He cites
General Motors Corp. v. Hopkins,
548 S.W.2d 344 (Tex.1977) (overruled in part,
Turner v. General Motors Corp.,
584 S.W.2d 844, 847 (Tex.1979)), as support. Dardis appears to characterize
Hopkins
as holding that the circumstantial evidence of a defect which arises from the malfunction can be
any defect
and does not have to be a defect in design, manufacture, or marketing. We find no support for that contention in
Hopkins
or elsewhere. In
Hopkins,
the plaintiffs entire case was tried on the theory of a design defect.
The portion of the opinion cited by Dar-dis was in the court’s discussion of General Motors’ claim of misuse. The relevant paragraph stated:
There is a related question of the proof of the condition of a product as of the time when the product left the hands of the defendant supplier. If the plaintiff has no evidence of a specific defect in the design or manufacture of the product, he may offer evidence of its malfunction as circumstantial proof of the product’s defect.
See, Williams v. Ford Motor Co.,
411 S.W.2d 443 (Mo.App.1966);
Greco v. Bucciconi Engineering Co.,
407 F.2d 87 (3rd Cir.1969). The age and use of that product during the time intervening between the purchase and malfunction will tend to support or defeat the circumstantial weight of the malfunction as proof of the original defect.
This statement in
Hopkins
must be reconciled with cases such as
Hernandez v. Nissan Motor Corp. in U.S.A.,
740 S.W.2d 894 (Tex.App.—El Paso 1987, writ denied), which held that “[t]he mere fact that an accident occurred is not sufficient proof that the automobile was defective.”
Id.
at 895. Similarly, in
Carroll v. Ford Motor Co.,
462 S.W.2d 57 (Tex.Civ.App.—Houston [14th Dist.] 1970, no writ), the court wrote, “[sjtrict liability eliminates both privity and negligence, but it does not prove the plaintiffs case. The plaintiff must prove that he has been injured or damaged by a defective product, and the mere possibility that this may have occurred is not enough.”
Id.
at 61.
See also Bass v. General Motors Corp.,
447 S.W.2d 443 (Tex.Civ.App.—Fort Worth 1968, writ ref'd n.r.e.).
The Texarkana Court of Appeals was faced with these apparently conflicting holdings in
Sipes v. General Motors Corp.,
946 S.W.2d 143 (Tex.App.—Texarkana 1997, writ denied). In that appeal from a summary judgment granted in favor of General Motors on Sipes’s strict products liability claim, the court held that where a plaintiff identifies a specific sealed mechanism which did not function as it was designed
and there was no evidence of tampering, that circumstantial evidence of a defect is sufficient to raise a question of fact on the issue of a manufacturing defect.
Id.
at 155-56. It next considered the issue o'f a design defect and found that General Motors’ summary judgment evidence of the purpose of the chosen design failed to affirmatively negate the existence of a practical safer alternative.
Id.
The court performed the same analysis with regard to the failure to warn. It analyzed this question even though it was not specifically alleged by the plaintiff. Athough
Sipes
was a summary judgment case, one teaching of that case appears clear; even when a plaintiff relies on the malfunction of a specific component as circumstantial evidence of a defect, that evidence must support one of the types of defects recognized
in strict products liability; design, manufacturing, or marketing.
Other cases in which a defect was found from circumstantial evidence are more similar to
Sipes
than the case before us. In
Darryl v. Ford Motor Co.,
440 S.W.2d 630 (Tex.1969), the plaintiff brought a strict products liability claim based on the failure of the brakes in his new truck. There, the plaintiff did not rely on the fact of the accident as proof of the defect, but produced a bent push rod from the brake system. Although the plaintiff was not required to establish the specific cause of the defect, like the plaintiff in
Sipes,
his evidence did establish, rather than merely infer, the failure of a specific component.
Unlike the undisputed fact in
Hopkins
that the vehicle air bag did not activate, whether the aircraft engine failed here was a heavily disputed issue at trial. The argument advanced by Dardis starts with the assumption that the engine failed. The only evidence of failure of the bleed valve is the inference made by Hamill that the bleed valve must have failed because he excluded the possibility of malfunction in the other engine systems. Dardis then uses this circumstantial evidence that the bleed valve faded to function properly as the circumstantial evidence that there was a defect of an unknown type. The same expert who opined that the valve failed stated that the failure was due to age and there was no defect in the design, manufacture, or marketing of the component.
In his motion for rehearing, Dar-dis contends that this analysis disapproves of circumstantial evidence to establish product defects. We disagree. Circumstantial evidence permits a factfinder to infer the fact to be proven by the circumstances shown by the proponent of the fact. Our rejection of the conclusion asserted by Dardis stems not from the circumstantial nature of the evidence, but the fact that in his argument, the conclusion of one inference was itself used as circumstantial evidence, requiring another inference to establish the ultimate fact. It is well established that facts may not be found based on chains of inferences.
See Hunsucker v. Omega Industries,
659 S.W.2d 692, 697 (Tex.App.—Dallas 1983, no writ).
The only theory under which we could find that the age of the valve could be a manufacturing defect is if Turbines were treated as the manufacturer of the engine. Dardis does not cite, and we have not found, any controlling authority holding that one who reassembles an engine may be treated as a manufacturer. On rehearing he cites three cases which, he contends, support the imposition of strict liability in this circumstance. Two of the three additional cases cited by Dardis concerned the liability of defendants who modified a product rather than simply performing mechanical work on it.
Green v. Los Angeles,
40 Cal.App.3d 819, 115 Cal.Rptr. 685 (1974) (crane);
Court v. Grzelinski,
72 Ill.2d 141, 19 Ill.Dec. 617, 379 N.E.2d 281 (1978) (automobile gas tank). Because Turbines did not modify the engine, those cases are inapplicable. The third case,
Crandell v. Larkin & Jones Appliance Co.,
334 N.W.2d 31 (S.D.1983), appears to offer some support for Dardis’s position. In
Crandell,
suit was filed against the appliance company that refurbished and sold a used clothes dryer. Due to the malfunction of its thermostats, which were pitted and worn, the dryer caused a fire. The court found the seller liable noting “[t]he application of strict liability to sellers of used products, who rebuild or recondition those products, helps to protect the reasonable expectations of consumers.”
Id.
at 34.
See also Markle v. Mulholland,
265 Or. 259, 509 P.2d 529 (1973) (retreaded tire).
However, the most nearly analogous cases we have found, also from other jurisdictions, support a contrary conclusion. In
Winans v. Rockwell International,
705 F.2d 1449 (5th Cir.1983), applying Louisiana law, the court held that strict liability would not apply to an engine manufacturer
for defects resulting from its rebuilding of an engine it originally manufactured, but would only apply for defects in the original manufacture.
Id.
at 1542. On rehearing, Dardis seeks to distinguish
Winans,
by noting that under Louisiana law of products liability, liability was limited to manufacturers and had not been applied to sellers of new or used products. However,
Winans
is representative of a series of cases declining to apply strict products liability to those whose relationship to a product is primarily performing service on, rather than creating, the product.
Win-ans
relied on
Johnson v. William C. Ellis & Sons Iron Works,
604 F.2d 950 (5th Cir.1979), which discussed policy reasons for not extending strict liability to those who merely perform work on products.
Id.
at 955.
Johnson,
in turn, cited the annotation
Application of Rule of Strict Liability in Tort to Persons Rendering Services,
29 A.L.R.3d 1425, 1970 WL 22208 (1970). That article, and in particular the supplement to it, recite numerous cases supporting the conclusion in
Winans. See, e.g., Hoffman v. Simplot Aviation, Inc.,
97 Idaho 32, 539 P.2d 584 (1975). Our review of these cases, and the additional cases cited by Dardis, convinces us that our disposition of this issue is correct.
Dardis concedes that Turbines’ conduct in assembling the engine “does not precisely fit a category of ‘manufacturing’ defect” because it did not manufacture the bleed valve. He argues, however, that when Turbines “rebuilt the engine, [it] failed to follow the manufacturer’s recommendation as to replacement of an old valve with a new or fresh valve.” While this argument might be relevant to a negligence claim, it is not relevant to Turbines’ liability in a strict products liability cause of action.
See Winans, supra.
Moreover, the argument is not supported by the evidence.
Each of the service bulletins relied on by Dardis appears in the record. As we noted in the factual statement, each sets out the specific steps required to implement the bulletin. An examination of the steps required to implement each bulletin reveals that none of them required replacement of the valve or the rubber diaphragm. In each case, the instructions required the valve assembly to be modified rather than replaced with a new part. This is made clear from the step in each set of instructions directing the mechanic to, for example, “[rjeidentify by striking out old compressor bleed valve assembly part number and marking new P/N 3010143 in an adjacent position.” Therefore, even if each service bulletin had been followed exactly, the compressor bleed valve would have a new part number, but not a new rubber diaphragm.
On this record, we are convinced that the evidence of a defect in the design, manufacture, or marketing of the engine was no more than a surmise or suspicion of a fact and amounted to no evidence, and that, as a matter of law, Turbines could not be liable for a manufacturing defect resulting from its performance of mechanical work on the engine. Therefore, we hold that submission of question one was error and sustain Turbines’ first issue. We note that our holding that there is no evidence of a defect in design, manufacturing, or marketing is not based on Turbines’ appellate argument that the trial court erred in failing to submit a question asking the jury to specify whether any defect was one of design, manufacturing or marketing. Turbines failed to present the trial court with the question it now contends should have been submitted. That failure precludes our consideration of that omission as a ground of reversal. Tex.R. Civ. P. 278. We also reject Turbines’ argument that the trial court erred in submitting question one because it had directed a verdict in favor of G & G on all claims, including one for strict products liability. Turbines has presented no authority in support of its argument that dismissal of a strict products liability claim against one defendant requires dismissal of that cause
of action against another defendant, or estops the plaintiff asserting those claims.
We next address Turbines’ third issue which assigns error to the submission of jury question two on negligence. The elements of a negligence cause of action are 1) a legal duty, 2) breach of that duty, and 3) damages proximately caused by that breach.
Greater Houston Transp. Co. v. Phillips,
801 S.W.2d 523, 525 (Tex.1990). Turbines does not challenge the existence of a duty. Its challenges are to the legal sufficiency of the evidence on breach and proximate cause.
Expert testimony is necessary when the alleged negligence is of such a nature as not to be within the experience of the layman.
Roark v. Allen,
633 S.W.2d 804, 809 (Tex.1982) (diagnosing a skull fracture);
Hager v. Romines,
913 S.W.2d 733, 734-35 (Tex.App.—Fort Worth 1995, no writ) (flying an airplane and aerially applying herbicide). The expert testimony must establish both the standard of care and the violation of that standard.
Id.
at 734-35. The performance of mechanical work on turbine aircraft engines is not within the experience of a layman. Therefore, Dardis was obligated to present expert testimony on the standard of care and that Turbines’ conduct did not meet that standard.
In a reply brief, Dardis argues that Turbines’ failure to object to the definition of ordinary care as “that degree of care that would be used by a person of ordinary prudence under the same or similar circumstances” waives any complaint as to the standard of care. He does not cite any authority for that proposition. We disagree. The circumstances at issue were those of an aircraft turbine engine mechanic. Because a lay jury could not know the degree of care of an ordinary person in those circumstances without expert testimony, Dardis was obligated to present evidence on that issue.
Dardis lists seven acts of alleged negligence with references to the record in support of the submission of question two. In our review of Turbines’ third issue, we must examine the evidence of each alleged act to determine if it provides legally sufficient evidence for a jury to find Turbines breached its duty. We address each alleged act in turn.
A. Turbines performed work for which they were not qualified to perform according to the manufacturer’s recommendations.
It is not clear from the use of the word “qualified” whether Dardis contends that Turbines did not possess the necessary skills to perform the work or they were not authorized to perform the work. The testimony cited in support of this contention is that of Don Hamill, who opined that Turbines’ repair station certificate prohibited it from performing work from the overhaul manual. He did not testify that Turbines lacked the skill to perform the work.
Hamill, who was not a licensed mechanic and had no training in the Federal Aviation Regulations, offered no reference to any regulation in support of his testimony that Turbines was not permitted to work from the overhaul manual. Even if his statement concerning use of the overhaul manual was accepted as true by the jury, he did not testify that a reasonably prudent mechanic working under that repair station certificate would not have done the work Turbines did. The only other theory on which Hamill’s testimony could support a finding of breach is that by violating this unidentified regulation, Turbines’ conduct was negligence
per se.
However, Dardis did not plead negligence
per se
or submit or obtain a jury issue on that question. There is no evidence that Turbines breached its duty by undertaking to perform the work it did on the engine.
B. Failing to obtain and review the engine history file before Turbines “rebuilt” the engine to determine the cause of the previous engine failures.
The evidence cited in support of this allegation shows that Turbines did not
determine the cause of the “loud noise” leading to the in-flight shutdown of the engine before it was sent to Pratt & Whitney. There was no evidence of the standard of care for an aviation mechanic or that the failure to make this determination fell below that standard of care. The closest evidence Dardis presented was his cross-examination of Haight, when he asked if it would be “a prudent thing” to get the full history of an engine that had suffered damage. Haight’s response that “[i]t’s always beneficial” cannot be taken as evidence of either the standard of care or that the failure to obtain that information was a breach of that standard.
C.Failing to replace the compressor bleed valve when rebuilding the engine as recommended by the manufacturer’s service bulletin.
The evidence in support of this contention fails in two independent respects. First, as discussed above, although they contained the word “replace,” none of the manufacturer’s service bulletins required replacement of the compressor bleed valve. As discussed above, each provided specific directions for modifying some portion of the valve and altering the part number to reflect the modified status of the valve. They did not direct replacement of the valve with a new part, although that was a permissible method to implement the service bulletins. Second, and most importantly, there was no evidence of the standard of care of an aircraft powerplant mechanic with regard to service bulletins or that the failure to implement bulletins was below that standard of care. Hamill explained the purpose and function of service bulletins and quoted from the bulletin that it was “recommended—desirable.” He did not state that a reasonable and prudent aviation mechanic would have implemented the bulletin or that the failure to do so fell below the standard of care of such a mechanic. Consequently, there is no evidence to support a finding that the failure to implement the service bulletins was a breach of Turbines’ duty.
D. Failing to fully inspect or test the compressor bleed valve.
The record is directly contrary to this contention. Haight testified on direct that the bleed valve was inspected as part of the hot section inspection and tested. He also described the specific steps performed in a functional test of the valve. Dardis appears to argue that the valve was not tested because the engine records do not contain a specific entry for that component. Haight explained that the engine build records certify performance of the steps in the engine manual and that the records do not list each component individually. The evidence shows that the valve was inspected and tested.
E. Failing to adequately inspect and repair the power section of the engine following Dardis’s first flight.
The only evidence cited in support of this contention is the testimony of Jim Mills of Turbines and the warranty repair invoice from Turbines. Mills testified that Dardis complained of sparks and elevated temperature in the engine. Turbines inspected the engine and found no problems, but replaced a bearing. Mills explained to Dardis that the -6 engine model operated approximately 100 degrees hotter than the -20 engine model. There was no evidence of Turbines’ standard of care with regard to inspection arid repair after such a complaint or that its conduct fell below that standard of care.
F. Failing to investigate the elevated engine temperature reported by Dardis.
As evidence of Turbines’ negligence, Dardis cites the testimony of Hamill that “[i]f you get a shift of two hundred degrees, it’s telling you something,” and “the increase in temperature seen by Mr. Van Dardis prior to his crash is indicative, in my opinion, that the bleed valve was saying I have got a little hole in it. [ ] The trouble shooting in the manual tells
you to check the bleed valve. That was not done.” However, Hamill admitted on cross-examination that the engine manual did
not
require any action if the temperature was below 995 degrees. The temperature reported by Dardis was 810 degrees. He also agreed that none of the engine manuals or service bulletins directed a mechanic to determine the cause of a temperature increase that was below 995 degrees. Hamill did not testify that an ordinary and prudent aviation mechanic would have disregarded the service manual and attempted to determine the cause of a temperature increase or that the failure to do so fell below the standard of care for an aviation mechanic.
G. Certified that the engine and power section were inspected, repaired and airworthy when they were not.
The only evidence cited in support of this argument is an invoice containing a certification of inspection and airworthiness. The term airworthy is defined in section 603(c) of the Federal Aviation Act (49 U.S.CApp. § 1423), and with regard to aircraft has been held to mean that the aircraft is 1) in conformity with the aircraft type certificate, and 2) in a condition for safe flight. We have found no authority that a different standard would apply to the airworthiness of aircraft components. Whether the engine was airworthy is a matter which required expert testimony. Dardis failed to provide expert testimony that the engine was not airworthy. Without such testimony, we may not simply infer lack of airworthiness from evidence of a failure. Moreover, there is no evidence of the standard of care for performing such an inspection or that Turbines’ inspection or certification were below that standard of care.
Because the conduct of Turbines was such that expert testimony was necessary to establish the standard of care and breach of that standard, after our examination of each of the acts alleged to have been a breach of its duty and the supporting evidence, we find that there was no legal evidence of breach. On rehearing, Dardis does not challenge our holding that there was failure of proof on the standard of care or breach, but rather, that those elements could be inferred under the doctrine of
res ipsa loquitur
based on the jury’s general knowledge. Although not cited by Dardis, a similar argument was accepted by our supreme court in
Harmon v. Sohio Pipeline Co.,
623 S.W.2d 314, 315 (Tex.1981). This holding makes our disposition of Turbines’ second issue challenging the submission of an instruction on
res ipsa loquitur
particularly significant, and ultimately dispositive of Dardis’s final argument on the third issue.
The instruction challenged by Turbines stating the doctrine of
res ipsa loquitur
provided:
In answering this question [on negligence] you may infer negligence by a person but are not compelled to do so, if you find that (1) the character of the occurrence is such that it would ordinarily not happen in the absence of negligence and (2) the instrumentality causing the occurrence was under the [sole] (sic) management and control of such person at the time that the negligence, if any, probably happened.
Turbines argues that the instruction was improper because “Dardis did not negate the possibility of pilot error as a cause of the accident” and that he did not establish that the instrumentality causing the accident was in Turbines’ control.
When applicable, the doctrine of
res ipsa loquitur
creates a permissible inference of negligence on the part of the defendant. The purpose of
res ipsa loqui-tur
is to relieve the plaintiff of the burden of proving a specific act of negligence by the defendant when it is impossible for the plaintiff to determine the sequence of events, or when the defendant has superior knowledge or means of information to determine the cause of the accident.
Jones v. Tarrant Utility Co.,
638 S.W.2d 862, 865
(Tex.1982). As correctly stated in the trial court’s instruction, this doctrine is applicable only when the following two factors are present: (1) the character of the injury is such that it would not have occurred in the absence of negligence; and (2) the instrumentality which caused the injury is shown to have been under the sole management and control of the defendant.
Gaulding v. Celotex Corp.,
772 S.W.2d 66, 68 (Tex.1989). The first factor is necessary to support the inference of negligence and the second to support the inference that the defendant was the negligent party.
Mobil Chemical Co. v. Bell,
517 S.W.2d 245, 251 (Tex.1974). Many jurisdictions recognize a requirement that the plaintiff establish the accident was not the result of any voluntary action or contribution by the plaintiff.
See, e.g., Newing v. Cheatham,
15 Cal.3d 351, 124 Cal.Rptr. 193, 540 P.2d 33 (1975). Although this requirement, derived from § 328D(l)(b) of the Restatement (Second) of Torts (1965), is not stated as an element by Texas courts, they do require that the plaintiff “so reduce” the likelihood of other causes that the jury can reasonably find by a preponderance of the evidence that the negligence, if any, lies at the defendant’s door.
Mobil,
517 S.W.2d at 251.
Turbines challenges both factors of the doctrine. It initially argues that the first factor was not met because “expert testimony is required unless the alleged negligent act is plainly within the common knowledge of laypersons” and cites the medical malpractice case of
Haddock v. Arnspiger,
793 S.W.2d 948, 951 (Tex.1990) as support. It then cites
McKinney v. Air Venture Corp.,
578 S.W.2d 849, 851 (Tex.Civ.App.—Fort Worth 1979, writ ref'd n.r.e.), for the proposition that flying an aircraft is not within the knowledge of the average person. These arguments fail to address the first element of
res ipsa loqui-tur.
We initially note the doctrine of
res ipsa loquitur
was not involved in
McKinney.
Next, the application of the doctrine to medical malpractice cases is subject to different standards than other types of negligence cases.
Haddock,
793 S.W.2d at 951. Finally, because application of the doctrine does not require ’evidence of a specific act of negligence, the focus is not on “the alleged negligent act” but on the nature of the accident or injury. General knowledge of aviation has progressed sufficiently that the average person can form an opinion as to whether it is unlikely that aircraft would crash in the absence of negligence.
Prosser and Keeton on Torts,
5th ed. § 39 (West 1984).
In challenge of the second factor of
res ipsa loquitur,
Turbines argues that it did not have the requisite control over the engine and that Dardis failed to negate his own negligence as a cause of the accident. Turbines does not contest that our review of a trial court’s decision to submit a jury instruction is governed by the abuse of discretion standard.
See State Farm Lloyds v. Nicolau,
951 S.W.2d 444, 451 (Tex.1997). In
Mobil,
the court wrote that “sound exercise of this discretion is particularly important in
res ipsa
cases in order to accommodate the wide range of situations which can arise in these cases.” 517 S.W.2d at 256.
In support of its argument that it did not have sufficient control to justify an inference of negligence on its part, Turbines points to evidence that it did not exercise any control over the compressor section of the engine, which included the bleed valve, since June 1993, almost a year before the accident. Dardis responds by citing a passage in
Mobil
that it “is sufficient if the defendant was in control at the time that the negligence inferable from the first factor probably occurred, so reasonable probabilities point to the defendant and support a reasonable inference that he was the negligent party.”
Id.
at 251. The difficulty in applying this holding to the facts of this case is that the negligence inferable from the first factor could be
either
negligence in assembling the engine or negligent operation of the aircraft by
Dardis. This uncertainty weighs against application of the doctrine on these facts.
Although there are no Texas cases addressing the control factor of
res ipsa lo-quitur
in similar situations, the available authority suggests that the second element was not satisfied. In
Marathon Oil Co. v. Sterner,
632 S.W.2d 571 (Tex.1982), the plaintiff was the employee of a contractor hired to perform work on a tank at a plant owned and operated by Marathon. He asserted the applicability of
res ipsa loquitur
when he was injured by exposure to gas in the tank. The court found the doctrine inapplicable on the issue of control because the contractor had been performing work on the tank for a week before the injury and it was “equally probable” that the negligence was attributable to the contractor.
Id.
at 574.
Opinions from other states support the view that, when
res ipsa loquitur
is applied to aircraft collision cases, it is the pilot who is in exclusive control.
Tompkins v. Northwestern Union Trust, Co.,
198 Mont. 170, 645 P.2d 402 (1982), is factually similar in that, as here, the parties presented two equally plausible theories of the crash, mechanical failure and pilot error. The court held that it was the pilot who was in control of the instrumentality even though another party was responsible for maintenance of the aircraft.
Id.
at 406. The fact that the pilot in
Tompkins
died and the only testimony on the crash was by conflicting expert opinions does not alter the analysis because here, Dardis’s testimony of mechanical failure was controverted by three pilot-witnesses, leaving two equally plausible theories.
See also Winans,
705 F.2d at 1455;
Webb v. Zurich Ins. Co.,
194 So.2d 436 (La.App.1966)
(res ipsa
could not be applied to manufacturer after delivery because of lack of control);
Shemwell v. Ailshire,
384 S.W.2d 104 (Mo.App.1964) (owner who rented aircraft not liable under
res ipsa
when engine failed on landing because pilot had control);
LeJeune v. Collard,
44 So.2d 504, 506 (La.App.1950) (where there was evidence of either mechanical failure or negligence by student pilot, for
res ipsa
purposes, phot was in control of plane, not flight school owner).
Turbines also argues that Dardis failed to show the applicability of
res ipsa loqui-tur
because he “failed to show that pilot error did not contribute to the accident.” It cites
Winans, supra,
and
Curry v. Chevron, USA
779 F.2d 272 (5th Cir.1985), for the proposition that in aviation cases, the plaintiff must establish that pilot error did not contribute to the accident. Dardis seeks to distinguish those cases on the basis that each involved a challenge to the denial of a
res ipsa loquitur
instruction rather than a challenge to the granting of an instruction, and that they were based on Louisiana law which, unlike Texas law, requires a plaintiff asserting
res ipsa lo-quitur
to totally negate other potential causes.
Although we agree with Dardis that the nature of the challenge to the trial court’s action is relevant, we do not agree with his characterization of Louisiana law. The Fifth Circuit did not require the plaintiffs to totally negate other potential causes, only to establish that the defendant’s negligence was the “most plausible” explanation for the accident.
Winans,
705 F.2d at 1454. We do not agree that this standard differs in any meaningful way from the application of
res ipsa loquitur
in Texas law. The portion of
Mobil
relied on by Turbines provides: “The possibility of other causes does not have to be completely eliminated, but their likelihood must be so reduced that the jury can reasonably find by a preponderance of the evidence that the negligence, if any, lies at the defendant’s door.”
Id.
at 251. This statement provides limited guidance for determining when the likelihood of other causes has been sufficiently reduced. However, our supreme court explicated this statement in
Marathon Oil Co. v. Sterner,
632 S.W.2d 571 (Tex.1982). There the court wrote: “When the plaintiffs evidence only shows it is equally probable that the negligence
was that of another, the court must direct the jury that [the] plaintiff has not proven his case.”
Id.
at 574.
See also Le Blanc, Inc. v. Gulf Bitulithic Co.,
412 S.W.2d 86, 96 (Tex.Civ.App.—Tyler 1967, writ ref'd n.r.e.);
City of Houston v. Church,
554 S.W.2d 242, 243-44 (Tex.Civ.App.—Houston [1st Dist.] 1977, writ ref'd n.r.e.). This formulation appears in the opinions of other jurisdictions to consider the issue.
See e.g., Knowlton v. Sandaker,
150 Mont. 438, 436 P.2d 98 (1968)
(res ipsa
submission improper when evidence showed two equally plausible explanations).
Here, there was facially credible evidence supporting both Turbines’ theory of pilot error and Dardis’s theory of engine failure. Viewing the evidence as a whole, the theories are equally probable and we cannot say that Dardis established that the negligence of Turbines was the most probable cause of the crash.
In summary of Turbines’ second issue, our review of the factors relevant to the doctrine of
res ipsa loquitur,
we find the trial court erred in submitting an instruction on that doctrine. First, there was no evidence that application of the doctrine was necessary to fulfill its purpose of relieving the burden on the plaintiff when he cannot determine the sequence of events, or when the defendant has superior knowledge or means of information to determine the cause of the accident. Second, Dardis did not establish Turbines’ exclusive control over the instrumentality causing the injury, and third, Dardis failed to show that negligence by Turbines, as opposed to his own negligence, was the most probable explanation of the crash. We therefore sustain Turbines’ second issue.
Our sustention of Turbines’ second issue also provides the resolution of its third issue on negligence. Because we hold that Dardis failed, as a matter of law, to establish the applicability of
res ipsa loquitur,
and presented no evidence that Turbines breached its duty to him, we sustain Turbines’ second and third issues and reverse the judgment in favor of Dardis on his negligence claim.
Turbines’ fourth issue challenges the sufficiency of the evidence supporting the jury’s finding of gross negligence. We need not address this issue because such issues are necessarily conditioned on a proper finding of negligence.
Ballesteros v. Jones,
985 S.W.2d 485, 500 (Tex.App.—San Antonio 1998, no pet. h.).
Because of our disposition of Turbines’ first three issues, we need not address its fifth issue challenging the admission of certain evidence. For the reasons we have expressed, we overrule the motion for rehearing and continue our rendition of judgment that appellees take nothing against Turbines.