The use of a torque wrench in the assembly of an engine, as well asother parts of a race car, is necessary for the correct tightening ofnuts and bolts. All threaded fasteners have a torque value. Materialsand designs are different, but the manufacturer has designated atightening torque for each.
Under-tightening does not provide proper thread loading. Over-tighteningcan stretch a fastener beyond its ability to recover. This weakening cancause the fastener to fail prematurely.
You should use a torque wrench to achieve the correct tightening torqueon any given fastener. But what if your torque wrench is inaccurate? Doyou have any idea if it is? If a torque wrench doesn't tell you thetruth, how do you properly assemble an engine?
The answer is that a torque wrench must be checked periodically.
One business that does exactly that is Calibration Specialities ofIrving, Texas. The decision-maker there is Rick Heck. During a recenttour of the company, Heck showed me the lab where they test, repair, andadjust torque wrenches, among other things. There were severalworkstations and devices for testing from the largest to the smallesttorque wrench. There were hundreds of pigeon holes filled with smallparts for any brand of torque wrench. I watched as a wrench wasdisassembled. A few small parts were carefully changed and it went backto the test bench where it then passed.
I asked Heck about the tolerance to which a wrench should be held. "Atolerance of less than +/- 4 percent for the older wrenches isconsidered acceptable," he said. "Newer-style wrenches run to about +/-3 percent. These are the accepted norms."
What about the frequency at which wrenches are checked? "The airlinessend theirs in about once a year," Heck said. "Some racing enginebuilders check theirs as often as every 30 days. They'll do it moreoften if a wrench has been dropped or mishandled. We check it and adjustits accuracy if necessary. The turnaround time is usually a week or so."
Heck says a torque wrench should be handled like the precision tool itreally is. "Tossing them in the toolbox and digging them outoccasionally is not the way to keep them accurate," he says.
"In order to properly test, we have to keep our equipment checked to alevel of about one half of one percent of dead true," Heck says. "Theprofessionals at the airlines and professional race engine builders allhave their wrenches checked and adjusted as necessary."
The materials needed to build...
The materials needed to build your own torque wrench tester. Only awelder and a metal cutting saw are needed for construction. Don't laughat the water bottle; it is a known weight.
1 piece of 1-in pipe, 8-ft long
2 pieces of 1-in square tubing,7-in long
1 piece of 1-in x 1-in flat bar or other material for a spacer
1 piece of 2-in angle, 5-in long
1 piece of 13/4-in tubing, 2-in long
1 5/8-in x 31/2-in bolt
3 5/8-in nuts
2 6202 ball bearings
1 package of J-B Weld
1 1/4-in or 1/2-in wide steel tape
1 gallon jug of water
25 pounds of balance weight
The 13/4-inch diameter tube...
The 13/4-inch diameter tube has already been welded to the angle. A7-inch long square tube is then welded on as shown.
The individual racer working in his home shop may not have $125 or so tospend for periodic calibration by a professional testing lab. But nowthere is a way to do it yourself. It requires about $30 worth ofmaterials and it can be built in a few hours.
The 5/8-inch bolt is shown...
The 5/8-inch bolt is shown with the two bearings and the 1-inch-longspacer. The spacer can be of any material. It only holds the bearingsapart until the J-B Weld is set.
I'll walk you through the process. You can build your torque wrenchchecker from whatever materials you have available. My materials list,which follows, contains things easily obtainable or plentiful in myshop.
Building the torque wrench checker is not difficult. In addition to thematerials, you will need a welder. You will need a way to cut steel--soeither a torch, band saw, cut-off wheel, or diamond-coated beaver teethwill do.
I purchased the 6202 ball bearings from a local auto parts store. Theywere selected because a 5/8-inch bolt is a close fit in the bore. Thereis no machine work required other than sawing parts to length. Followthe photos for the building sequence.
Once complete, attach the balance weight to the short end of the beam.You will need about 25 pounds. I used a set of clamp-on race car weightsfrom A&A Manufacturing. Move the ballast until the beam comes in contactwith the attached steel tape. This balance is critical to the accuracyof the tester.
Now suspend the water bottle on one of the marks on the beam. I usuallytest at 3 feet and 6 feet to get a low and high check. With the end ofthe steel tape extended under the beam, attach the appropriate socketand place the wrench on the end of the pivot bolt. With a slow, evenpull, the same way you would when tightening a bolt, apply pressure tothe wrench. Watch or feel when the wrench reaches a specific torque. Ifyou are testing at 3 feet and the water weighs 8 pounds, then the beamshould lift when the wrench reads 24 pounds. You will hear the steeltape snap back when the beam lifts off. Higher or lower readingsindicate the amount of error.
The bearings are coated on...
The bearings are coated on the outside with J-B Weld and slipped intothe tube. Don't let the J-B Weld run down on the bolt. Leave the bolt inplace to keep the bearings located. Clean up wet J-B Weld or otherepoxies with alcohol.
The last 7-inch piece of square...
The last 7-inch piece of square tubing is welded across the end of thefirst. It should be about 3/8-inch above the surface of the angle so thebeam will lay flat on it.
This is the heart of the accuracy...
This is the heart of the accuracy of this unit. Purchase a gallon waterjug at the grocery store. Take it to the meat department. Ask them toweigh it and print out a label with the exact weight. Multiply thatnumber times the number of feet from the pivot bolt. The answer will bethe foot-pounds of torque required to lift the beam.
This is the finished working...
This is the finished working section of the checker. The angle isclamped in a vise. The bolt is ready to be screwed into the nut on thebeam. A steel tape is tie wrapped to the tubing. This shows the way itshould be extended when ready for use. When the beam rises, the tapewill audibly click back against the tube.
The 8-foot-long beam has a...
The 8-foot-long beam has a 5/8-inch nut welded to it 18-inch from oneend. This will attach to the pivot bolt. A lock nut must be used againstthe beam nut. The photo shows the layout of the beam marks. For maximumaccuracy, start with the 12-inch mark at the center of the bolt. Thenmark off each foot toward the long end.
Only at a professional testing facility such as Calibration Specialtiescan a torque wrench be adjusted to correct inaccuracy. However, for theindividual racer who knows a wrench is off by say 9 percent will be ableto compensate for that error by adding or subtracting that amount.
Results Of Sample Torque Wrenches Checked
*Acceptable error +/- 4%
|Wrench Type||CORRECT TORQUE||CHECKED TORQUE||ERROR|
|1/2" Drive Beam Type||Low 26 ft.-lb./High 61 ft.-lb||28 ft.-lb./64 ft.-lb.||7.7%/4.9%|
|1/2" Driv Clicker Type||Low 26 ft.-lb./High 61 ft.-lb||29 ft.-lb./66 ft.-lb.||11.5%/8.2%|
|3/8" Drive Beam Type||313 in.-lb.||328 in.-lb.||4.8%|
I tested three of my own torque wrenches: a 1/2-inch drive beam type, a3/8-inch drive beam type, and a 1/2-inch drive "clicker" type. Theresults are in the chart below. Remember +/- 4 percent is the outerlimit of acceptable error. I suggest checking your wrench at the valuesit will be used, for instance at the recommended torque for rod bolts,main bolts, and head bolts. Use a calculator to predetermine the amountof error to add or subtract. Then use the corrected number whentightening fasteners.
Loose race cars may be fast but tight bolts keep them that way!