E U Q R R O T ERTE V N P O U C OCK L l tteeeel ee s s ssee nnggiinn o o h h ee tt tthhee ee llllyy d d i i iinnss eeeenn rreeffuu dd aa n n o o eettw w ?? CCaa d aann eess s s e e b b oo eedd nn fflluuiid m aakk g o g o i i t m s t t s t l a eess tt.. W miiss iissssiioonn ttiim Whhaellss bbool nnssm m ee leevvaann e ttrraa nssm g m g m a m a oo l n bb nndd aa edd ttrraathhaatt ss gg iirrrree t hhiinn S sshhe iissm m D DS o o l O l O ss haann hoollee tt O O W W cch E wh E O O m mee tthhee w J J B BYY

The author wishes to thank Mike Brewer, Instructor at the Automotive Training Center in Exton, PA, for use of the split torque converters shown in this article.

38

May 1999

F

or many years, stick-shift vehicles got measurably better fuel economy and range than those with automatics. It was clear enough why: Oldstyle torque converters necessarily allowed an rpm loss between the crankshaft and the input shaft of the gearbox. When people called an automatic a “slushbox,” they weren’t far wrong. The slushy connection between the drive and driven elements was entirely through the PRNDLjuice. In this article we’ll look at how torque converters work, how they fail and how you can tell the difference. Let’s define our terms. The torque converter is a large donut-shaped (“toroidal”) coupling between the flexplate and the transmission input shaft. It bolts directly to the flexplate, which in turn bolts directly to the crankshaft. For balancing reasons, the bolt pattern between the torque converter and the flexplate often—but not always—allows only one orientation of one to the other, ensuring the right clock-position between flange and converter. The back of the torque converter ends in a notched or slotted tube that engages the center driving element of the transmission oil pump. So when the crankshaft turns, the transmission oil pump turns, squeezing PRNDL-juice through the hydraulic circuits and building pressure to work the valves, actuators and clutches.

Photos: Joe Woods

Antique Answers & Hot Oil

When you explain to an apprentice how an ignition system works, don’t you find that you usually describe first the workings of the now-obsolete contact point ignition system? From it, it’s much easier to explain and understand how solid-state and computer-controlled spark systems work. In that same way, if we look at the first “torque converters”—fluid couplings—the rest will be clearer. Some shop manuals for German cars still refer to the torque converter as the “Foettinger coupling,” after its inventor. These first ancestors of the torque converter replaced the clutch on WWI armored vehicles. The fluid coupling is effectively two paddlewheels in the same oil-filled drum—one part of the case, one turning freely, as you can see in the photo on page 40. There is always rpm slip between the impeller and the turbine of a fluid coupling, but the rpm stays roughly con-

stant, so as you raise the engine rpm you reduce the percentage lost as slip. At idle, that can be 100%; at full power, it probably drops to 20% or so, although the slip rpm is the same or slightly greater. Of course, all the energy that doesn’t go from the impeller to the turbine as mechanical torque becomes turbulence and heat. This is better than burning up clutch disks because it’s easier to change oil than disks. But not a lot better.

Today’s Toroidal Troubles The most important part of the torque converter is the fluid itself. Filling all the donut-shaped empty space in the converter, the oil receives 100% of the engine’s output torque and delivers the power to the transmission, minus the commission it deducts as waste heat. The complex pattern of oil flow in this toroidal (donut-shaped) chamber is the key to understanding the converter. The active role of the fluid in the converter, in fact, is the reason you measure transmission fluid with the engine running and at operating temperature: You want the torque converter completely filled when you check the dipstick. Once the engine stops, the converter gradually drains about half its oil back into the sump. Sealed inside the torque converter are several major components. The impeller has radial vanes welded in place on the rear half of the donut-shaped housing. As the crankshaft rotates the housing, the impeller’s vanes spin and carry with them the transmission fluid, filling the converter and throwing the fluid outward as well as spinning it. Older converters used axial impeller vanes like the fluid couplings; newer converters’ impeller vanes curve forward, in the direction of crankshaft rotation, to spin the fluid slightly faster at the perimeter than the converter shell moves itself. One of the curiosities of the modern torque converter, if you look closely, is that the vanes go the wrong way. Compare the curvature of a torque converter’s vanes to those of a water pump impeller; they curve the opposite way. Doesn’t this reduce the torque converter’s capacity to pump oil? It sure does—but that’s the point. A torque converter does not work like a basement sump pump, to move as much fluid as possible as quickly as possible. It uses the oil only as an energy transfer coupling. Moving it around in the transmission and torque converter is friction, a necessary evil, not the point of the game. The torque converter’s vanes are shaped to maximize the speed and momentum of the oil at the outside edge, to maximize the inertia transmitted from the im-

May 1999

39

TORQUE CONVERTER LOCKUP peller to the turbine. As it moves through the impeller, the oil moves outward and winds forward, in the direction of crankshaft rotation. The oil next strikes the turbine vanes, which are equal in size and shape to those of the impeller but curved in mirror-image, to capture and slow the spinning fluid, extracting as much whirling torque as possible. The turbine is not mechanically connected to the impeller (in the absence of an applied lockup clutch), but spins entirely from the inertia of the whirling oil. The turbine shaft splines directly to the input shaft of the transmission, turning whichever clutches and gears are hydraulically engaged in the transmission itself. As the oil

Then the sprag’s one-way clutch allows it to turn freely with the whirling fluid of the impeller and turbine. As we all learned in Auto Trans 101, a two-way freewheeling sprag will result in a vehicle with poor acceleration but satisfactory cruise speed; a sprag clutch seized on its one-way clutch, on the other hand, allows satisfactory early acceleration, but prevents normal cruise speed (and builds up transmission heat rapidly and dramatically, bluing the converter shell and cooking the clutches). Of course, once a torque converter is out and on the bench, you can directly test the sprag clutch by trying to rotate it in opposite directions with a suitable shaft. Once the stator is

oil, an arrangement possible only on very small vehicles with relatively unpowerful engines...and not very successful even then. Routing the oil through the transmission cooler, of course, can effectively shed this heat, but the energy and thus the fuel are wasted. If there were some way to lock the crankshaft to the transmission input shaft, in the same one-to-one ratio as in a manual transmission, the energy the engine produces as torque could be delivered at no discount to the working gears of the transmission. This is not something impossible. Centrifugal clutches on chainsaws and go-karts manage it, but they’d hardly do for the load and power of an automobile, al-

Original torque converters were more properly fluid couplings, connecting the driving impeller from the crankshaft and the driven turbine to the transmission input shaft only through fluid turbulence. Efficiency was not their long suit; they were used only to replace a friction clutch.

moves through the turbine, it spirals inward and against the direction of crankshaft rotation, dumping its torque against the turbine blades. As the fluid is driven in toward the center by its reaction against the turbine blades, it strikes the vanes of the stator. At comparatively low speeds, the stator locks against the hydraulic force that would turn it backward by a one-way sprag clutch on the tube extending from the transmission front pump housing. Once the speed of the engine and converter is high enough, the oil from the turbine is no longer moving backward with respect to the stationary front tube, so the force against the stator goes away.

40

May 1999

turning freely with the other elements, the fluid just passes between its vanes to reenter those of the impeller.

Cooling Conundrums The major design problem with this type of torque converter, obviously, is that there’s always a certain amount of fluid slip between the impeller and the turbine. This causes shearing turbulence in the oil, and such forcefully churned turbulence is, plain and simple, heat. Early automatic transmissions on some European cars employed air vanes or scoops welded to the outside of the torque converter to try to cool the

The impeller vanes are welded to the housing of the torque converter. Here we’re looking from the engine toward the transmission. As the crankshaf t turns, spinning the torque converter housing clockwise, the vanes scoop up the oil and spin it outward, increasing its rotational speed where it exits the impeller to somewhat above the speed of the shell. The dished-out circular trough in the middle of the vanes serves to isolate the fluid at the center of rotation, which would otherwise just churn around uselessly, making heat.

TORQUE CONVERTER LOCKUP A POUR-IN FIX?

I

’ve always suspected that nothing you pour in or spray on can possibly have any good effects on a car other than dissolving varnish. There are some clear exceptions, though. Some penetrating fluids really do help work a rust-seized bolt loose; some antiseize compounds really do keep spark plug threads and brake bleeders from practically welding themselves to their internal casting threads. At least one electrical contact spray I use really does work, and some good threadlocking compounds will prevent a nut or bolt from rattling loose. Maybe Lubegard is a transmission additive that should join this short list. Many independent transmission shops use the stuff routinely in their rebuilds. Lubegard claims dramatic improvements in the heat range of ATF when mixed with the additive. Claims are one thing, but a surprising number of carmakers have issued TSBs recommending the product for specific slipping and chattering problems, including those for torque converter clutches. Contact International Lubricants either through www.lubegard.com or at 800-333-LUBE (5823). They’ll tell you more than you ever knew there was to know about the properties, desired and undesired, of PRNDLjuice. Of course, Lubegard can’t repair an element that has burned off its friction surfaces, but many people who know what they’re doing believe it can extend the life of a transmission.—J.W.

though there were some automatic transmissions earlier that did use a centrifugal lockup. The lockup clutch we see today works in a more complicated but more satisfactory manner. Hydraulic pressure locks the turbine to the front of the torque converter case during those driving conditions when we want the crankshaft and the input shaft turning exactly together—steady cruise at highway speeds, for instance. Since either the valve body or the electronic “decision” circuits of the ECM determine engagement, the lockup clutch either definitely engages or definitely releases with minimal slip time. Once the lockup clutch engages, fluid in the

42

May 1999

converter just rotates at the same speed as everything else inside, with little or no heat-producing turbulence. The only fluid movement comes from the gradual circulation to keep cycling it through the cooler. The lockup mechanism—usually a frictional clutch operated by a separate hydraulic pressure circuit, sometimes a splined fitting similarly activated—is an additional element to the torque converter beyond the impeller, turbine and stator. Once it’s engaged, everything inside the torque converter serves only as rotating flywheel mass. An advantage of this is that it reduces the cooling load on the radiator and saves fuel as well. There are basically three ways a torque converter lockup clutch can fail: It can stay locked up, stalling the engine when the car is braked to a stop; it can never lock up, which shows up as an increase in fuel consumption and radiator temperature; or it can slip when engaged, allowing engine speed surges at a constant vehicle speed. How can you tell if a lockup clutch engages or not? For many purposes, you can judge transmission shift timing and quality subjectively, by the seat of your pants. But on many lockup converters, the engagement is so gentle and the change of engine speed so minor that it’s very difficult to tell without an auxiliary tachometer (except for those cars combining an automatic transmission with a dashboard tach). Virtually all torque converters will unlock if you apply the brakes or release the throttle. This disengagement is often easier to notice than the engagement, which is gradual on some cars and “vague” on others like Cadillac, where some lockup clutches use a viscous disk functioning rather like the multidisk and silicone-fluid interaxle “differential” on smaller allwheel-drive cars.

Thermal Controls With a lockup clutch in the torque converter, engineers can use an impeller and turbine with slightly less fluid-coupling between them, allowing more slip, more torque multiplication and, as an unintended side effect, more heat. Turbulence in the torque

The stator, perhaps the most unique element in the torque converter, is what enables the converter to multiply torque at lower rpm. At low engine speeds, the sprag clutch (center) locks the stator (seen here from the engine side). The stator vanes, small airfoils like turbine blades adjacent to the hub of the shafts, then reverse the direction of fluid flow and drive the f luid clockwise as seen from the crankshaft. Once engine speed passes a certain threshold, the turbine is turning fast enough that it no longer throws fluid “backward,” and the stator sprag clutch unlocks. After that, it’s just along for the ride.

converter is almost the only heat source in an automatic transmission, other than the momentary temperature buildup as the clutches engage and disengage. As a result, a number of automatic transmissions and transaxles will engage the torque converter lockup in manually selected lower gears to keep the transmission from overheating during extended steep hill climbs or while towing. Some control systems will engage the lockup clutch in most gears above a certain speed if the transmission gets too hot. The point of these comments is that you have to know the control system’s strategy for converter clutch opera-

tion, or you won’t know whether the applications and releases are working the way they’re supposed to. If the lockup clutch never engages or stays locked up until it stalls a braking vehicle, that’s another matter. But if it is just applying and releasing at what seem like odd moments, better find out what the program is before you start doing any “fixing.” Most lockup torque converter problems arise outside the torque converter, either in the valve body or in the electronic controls or engagement solenoid. Of course, if there are loud noises from inside the drum or acceleration problems characteristic of a sticking or freewheeling stator, that’s another story. But most of the work can be done without removing the transmission from the vehicle, once you know what the problem is. As with most work on today’s vehicles, careful diagnosis is the major portion of the task, and understanding the operation of the mechanism is key to the diagnosis.

Flushing the Converter While practically no vehicle manufacturer authorizes it, probably every transmission shop uses the technique of flushing the converter—typically by disconnecting the coolant line and pouring transmission fluid in at the same rate it comes out until it turns red. There are obvious risks to doing the job this way: If you pour too slowly,

The lockup clutch, evident here by the friction surfaces between the front of the turbine and the torque converter front housing, directly locks the engine’s crankshaft to the transmission’s input shaft, achieving the same oneto-one ratio as in a standard transmission. A few transmissions use a sliding spline to achieve the same result, while others use a viscous coupling to filter out slight engine torque pulses. Once the lockup clutch engages, the fluid in the converter effectively stops all relative motion between the impeller and the turbine, except for the gradual flushing of oil through the cooler lines.

the pump loses its prime and runs dry; if you pour too quickly, the level rises too much and you might get aeration of the oil. However, in the real world, where customers are sensitive to prices, and where it requires removal of the transaxle to get the converter out to flush with specialized tools, this is a technique most shops can master without disaster. Another option is to flush the trans-

BLUEING & BALLOONING

A

torque converter may look vaguely like a bagel, but its dimensions are not so tolerant. While built to exacting standards, there are two forces that can distort the shell, sometimes permanently and destructively—heat and pressure. Heat builds up naturally from the fluid shear, but ordinarily it should exit through the coolant lines and sink through the radiator into the cooling air stream. Let the radiator run low enough that the transmission heat exchanger is not in the cooled liquid, or let the coolant lines kink, or let the stator lock on the pump cover sleeve and the shell of the converter gets very hot—hot enough to blue the metal like a gun barrel. But this is thick sheet metal, not ordnance steel; if it gets hot enough long enough, it gets softer and bends. Ordinarily, the pressure in a torque converter is not that high, even when the lockup clutch engages. But couple a little too much pressure with enough heat to blue the shell metal, and you can find a converter that “balloons,” or gets thicker at the centerline. The first effect (which can come from excess pressure alone) is accelerated wear of the crankshaft’s thrust bearing, undetectable by ordinary pry-bar to-and-fro tests, since the steel “balloon” doesn’t deflate. Eventually it can start to rub metal to metal at the rear main or on the transmission front cover. That’s the end of the road for that converter. Remember, of course, you have to discover and correct the source of the excess heat and pressure, or you’ll be doing the job again...for free.—J.W.

mission with a machine designed specifically for the purpose. Several transmission flushers are out there, and they do a good job of removing contaminants from the transmission, converter and cooler. Some hook directly to the lines, others to the filter housing. Ask for a demo before deciding on what’s right for your shop. Regardless of the method, flushing will not tell you whether there has been any damage to the friction surfaces of the lockup clutch. In fact, that’s a conclusion you can reach only indirectly, by observing the operation of the converter when the car is running at cruise speeds. Dissections of returned torque converters have shown very few with a clutch that has worn out, even on very high-mileage units. Failure of the stator one-way clutch, a rare enough event, is far more common than a burned-out lockup clutch. For a free copy of this article, write to: Fulfillment Dept., MOTOR Magazine, 5600 Crooks Rd., Troy, MI 48098. Additional copies are $2 each. Send check or money order.

May 1999

43