The first in-depth study of different double-stack railcar designs has found disparities in ride qualities between different types of stack equipment.

The study, which will not be made public until Oct. 19, raises industrywide questions about ride quality of some double-stack equipment, particularly vehicles known as stand-alone cars with a single freight-carrying well.The study is based on tests commissioned by the Association of American Railroads, Burlington Northern Railroad and TTX Co. With double-stack trains touted by intermodal vendors as a major factor in reducing freight damage, the studies findings may affect shippers decision to use intermodal and in determining what kind of cars the carriers will purchase.

The study's authors stressed that test results are in line with their expectations, given the differences in the car types tested. Several industry officials who were co-authors of the study cautioned that test conditions did not mirror typical train operations.

"We would have expected and the test showed that stand-alone (single unit) cars did not do as well in terms of slack action as a five-platform car," said Keith Kieres, superintendent of mechanical engineering for Burlington Northern.

Slack action is the force exerted as freight-car couplers bump into each other when trains change speeds or go up or down grades.

The authors would not release specific results, saying that actual performance would be disclosed in a report to be given at an American Society of Mechanical Engineers meeting in Pueblo, Colo.

The tests covered four types of double-stack equipment:

* Cars with five wells, the original stack car and the only one that was extensively tested previously.

* Sets with three platforms connected by a drawbar that reduces longitudinal forces that influence ride quality.

* Single-unit stack cars with reduced slack couplers.

* Single-unit cars with conventional couplers.

"I think a lot of the issue boils down to what's considered a negative reflection," said Dick Dawson, chief engineer for TTX. "This was a comparative test of complete trains of double-stack cars with different types of car-body connection systems. This is a case in which comparisons don't lend themselves to a black-and-white evaluation."

"The question that is not directly addressed is what is acceptable performance," he said. "I'm not aware of any clearly established criteria of what is acceptable for merchandise traffic."

"This is not intended to be representative of the way trains would be run," Mr. Kieres said. "We had to test a train of all similarly equipped cars in order to get valid test results on the performance of each car type in those trains."

In practice, stack trains often have a mixture of differently equipped cars, including stand-alone, five-platform and other types.

Mr. Dawson also cautioned that the test measured only longitudinal forces, such as coupling, and did not consider other in-train physical forces that would not differ by car type. Rail freight shipments also experience side-to- side, or lateral, and vertical forces that were not part of the test.

Mr. Dawson, Mr. Kieres and Firdausi Irani, an official in the engineering department of the Association of American Railroads, said the study did show that the single-unit stack cars with standard couplers had better ride quality, at least for longitudinal forces, than the conventional 89-foot flatcars that once were the core of the intermodal fleet.

Today, about two-thirds of the intermodal capacity is articulated - or jointed - and/or double-stacks.

"Our analysis is that the tests showed exactly what we expected," said Bruce Harmon, a spokesman for Gunderson Inc., a manufacturer of double-stack equipment. "The important thing is that the single-unit double-stack car behaved extremely well compared to other stand-alone, conventional freight cars."

The study also tested performance of single-level, articulated cars, commonly called spine cars, but Mr. Kieres said instrument problems invalidated the results.