ECR Engines' Dyno Adapter System

A Huge Reliability Gain for the ECR Active Dyno Cell

The testing of the performance characteristics of race engines has become increasingly more complex and sophisticated in recent years, especially at the top levels of motorsport.

In the not-too-distant past, the engine testing program for many racers consisted of a few "pulls" on a passive engine dynamometer ("dyno"). A "passive" dyno is one which basically consists of (a) a power absorber (water brake, hydraulic brake, eddy current brake, etc.) and (b) some rudimentary form of control system which varies the resistance of the absorber so as to just match the torque being produced by the engine (for a steady-rpm run) or which varies the absorber load so as to allow a preprogrammed engine acceleration rate (100 rpm/sec, 200 rpm/sec, etc) over the span of a preprogrammed rpm range (known colloquially as a "pull").

An "active" dyno is a highly sophisticated system in which the absorber is typically a motor-generator with a permanent-magnet-rotor, controlled by a highly capable, programmable computer-based system. This type of dyno can (a) be driven by the engine and measure engine torque, (b) drive the engine and measure the driving torque, and (c) do combinations of those operations according to complex preprogrammed schedules. Using those capabilities, an active dyno can accurately simulate the actual usage which a race engine experiences throughout practice, qualifying and running an entire race.

Contemporary top level teams in motorsport (NASCAR, IRL, F1, Moto-GP, etc) use very sophisticated, environmentally-controlled engine test cells equipped with active dynamometers. An example of such a facility is the active test cell at ECR (Earnhardt-Childress Racing) Engines pictured below.

ECR Engines is an independent performance engine development and manufacturing company based in Welcome, NC, and is the sole supplier of race engines for all the Richard Childress Racing NASCAR teams, Earnhardt-Ganassi Racing NASCAR teams, and other notable teams.

In 2010, ECR Engines won ten NASCAR Sprint Cup races, six with RCR (Richard Childress Racing): {Bud Shootout, Talladega-1, Daytona-2, Michigan-2, Loudon-2, Talladega-2} and four with EGR {Daytona-1, Brickyard 400, Watkins Glen, Charlotte-2}.

In 2011, ECR engines won eight Sprint Cup races, seven with RCR (Daytona Duel-2, Fontana, Martinsville-1, Charlotte-1, Brickyard-400, Richmond-2, Talladega-2), and one with Furniture Row {DarlingtonSouthern 500}.

In both 2010 and 2011, ECR engines have won more than 50 races. In 2011, those wins include 23 victories in the top three NASCAR series, as well as 8 wins in the ARCA Racing Series presented by Menards, 5 wins in the Rolex Grand-Am Daytona Prototype Series, 2 victories in SCCA Trans-Am, as well as propelling RCR’s Austin Dillon to the 2011 NASCAR Camping World Truck Series championship and RCR's Ty Dillon to the 2011 ARCA Racing Series championship.

Active AVL Dyno at ECR Engines

Active Test Cell Control Room at ECR Engines

Winning race engines must be optimized to the very specific challenges presented by different race tracks. Here are examples of three very different race scenarios.

One lap around Talladega in a NASCAR Cup car is simply a full-throttle, never-lift cycle in which the engine accelerates from 8600 rpm exiting a corner to 8900 rpm at the entry to the next corner, then, still at full throttle, decelerates slowly back to 8600 as the cornering loads reduce the speed of the car.

By comparison, a lap around Martinsville consists of exiting the corners at part-throttle around 5800 rpm, then full throttle acceleration to 9500, then closed throttle and heavy braking back to 5800 on entry to the next corner.

On a road course (Watkins Glen, for example), the engine is subjected to even greater transients, exiting some corners at the low end of the usable torque range in a low gear, accelerating to max rpm through the gears and the near-instantaneous change in engine speed and the near-shock-load of shifting and clutch re-engagement. Corner entries can vary from a brief lift of the throttle followed by full throttle again, to the extreme of several downshifts and clutch-releases under braking, followed by full throttle acceleration on exit.

An active dyno can simulate an entire race at any track for which the team has gathered the appropriate data. Using data tables which quantify throttle position, engine acceleration / deceleration rates and shift / braking points for a given race course, an active dyno system can (a) vary the throttle settings, (b) apply the correct absorber resistance to allow the engine to achieve a scheduled acceleration rate, (c) suddenly change the absorber load to simulate a full-throttle shift and clutch release, (d) suddenly accelerate the engine under closed-throttle to simulate a downshift, and (e) drive the engine so as to achieve a scheduled off-throttle or partial-throttle deceleration, in programmable combinations and sequences so as to completely simulate laps around a given course.

From that brief discussion, it should be evident how critical the active test cell is to the success of a top-level engine developer, to insure by exhaustive testing that these amazing, 850+ HP 2-valve, pushrod, carbureted V8's can perform at up to 9500 RPM for up to 600 miles PLUS practice and qualifying. It is also evident that the hardware in an active cell must be able to survive a lot of abuse, day after day after day.

ECR Engines' active test cell uses a sophisticated AVL dynamometer system, as pictured above. The AVL absorber input components consist of a special driveshaft and two complex flexure-joints (pictured below). The flexures compensate for some of the axial growth of the shaft with rising temperature, but they allow for very little (a few thousandths) axial or angular misalignment.

The AVL-specified alignment procedure which is required to assure angular and axial alignment of the shaft and flexures is both demanding and time consuming, and precludes attaching an engine directly to the flexure joints.

AVL Dyno Input System

In order to allow rapid installation and removal of test engines, ECR Engines acquired an adapter which could be fixed rigidly in place relative to the AVL absorber, thereby avoiding the need to disturb the critical flexure alignment during engine installation and removal, while at the same time extending the distance between the absorber and the engine so as to allow the use of the complete race-car exhaust system in the cell.

The adapter consisted of a steel housing mounted rigidly to the floating cell floor to which the AVL absorber is mounted. The housing contained a 3-inch diameter shaft having an 8.5-inch diameter flange at the dyno-end to attach to the AVL flexure, and a beefy external spline (2-inch OD, 62 HRc Maraging-300) at the engine end. A modified bellhousing bolts to this adapter housing and provides the interface to which the engine attaches. The coupling between the engine crankshaft and the adapter was accomplished by a primitive "torsional absorber" which relied on the elastic properties of a few internal rubber bits, and which mated with the huge spline on the adapter shaft. This system is pictured below (left side of picture).

AVL Dyno Adapter

The adapter proved to be extremely unreliable to the extent that it was severely compromising the availability of the active dyno cell, thereby compromising the progress of several critical engine development programs. In as little as 3 hours of simulation testing, the adapter would experience one or more failures including: (a) destruction of the bearings which carry the adapter shaft, (b) resultant damage to the adapter shaft and / or housing; (c) destruction of the rubber-bits in the "torsional absorber"; (d) severe fretting of the massive 2" drive spline and mating female spline, contaminating the entire adapter area with abrasive iron-oxide dust.

Early in 2010, ECR Engines contacted EPI, Inc. in Amboy, Washington, to discuss potential solutions for these problems. EPI did some analysis of the existing system and identified five major design flaws. One of those flaws was the existence of a system first-mode resonant frequency at approximately 2500 engine rpm (assuming 4th order excitation) which was the suspected cause of a host of smaller problems.

Next, EPI designed a new system which would solve all 5 of the identified problems. The proposed system would use an adaptation of a torsional isolation system which they have used successfully in aircraft engine applications, coupled with a highly-modified version of the original housing, a highly-modified version of the original 3-inch internal adapter shaft, a modified bellhousing, and a completely new bearing system.

They presented this system design and the supporting calculations to the ECR Engines engineering staff and, after some discussion, ECR asked them to complete the design AND to build a prototype of the new system.

In early June 2010, EPI delivered the new adapter system to the ECR facility, and EPI's chief engineer came here to assist with the installation and testing. The first time an engine was fired on the new system, everyone commented on how much more quiet the environment was compared to all the hammering and clattering the old system produced. One ECR test engineer commented that he could now hear engine noises that had never been possible to hear before.

The EPI system has apparently solved the reliability problems. The new system (a) moved the problematic resonant frequency down to approximately 700 engine rpm, (b) solved the bearing failure problems, (c) eliminated the primitive rubber absorber, and (d) nearly eliminates the spline fretting. The active cell has been in nearly-constant use with this system since early June, 2010 and, as of this report (November 2010) has survived almost 200 million stress cycles and at least one engine explosion.

As part of this project, EPI, Inc. provided ECR Engines with a complete set of 3D-CAD models, engineering drawings and process sheets for every component in the system, as well as a detailed, illustrated assembly / disassembly manual. ECR and EPI have worked together to establish a "normal maintenance" schedule, and EPI has, at ECR's request, made modifications to the system to accomodate other engine packages and instrumentation systems, and is working on improvements to reduce the small amount of wear which the new system has seen in nearly 200 million stress cycles.

copyright © 2008 ECR Engines -- all rights reserved