Door closing force: what the spring gauge really means

Every few months someone sends me a photo of a spring gauge hooked to a car door with a reading of 120 N and a message saying the door is fine. It is fine, on that check. Then you watch it close and the last 200 mm come in like a truck, and you understand why the passenger complained. Two limits apply to a power-operated door, and inspectors check them separately. The closing force at the leading edge must not exceed 150 N, about the pull a 15 kg weight puts on a rope. The kinetic energy of the moving door must not exceed 10 J. Force is what the door pushes with when it is stopped. Energy is what it carries when it is moving, and energy is what hurts. The arithmetic is worth keeping in your head. Kinetic energy is half the mass times the square of the speed. A 60 kg door panel closing at 0.5 m/s carries 7.5 J. The same panel at 0.6 m/s carries 10.8 J, which is over the limit even though the motor torque has not changed at all. Speed costs you energy squared. This is why a door can read 130 N on the gauge and still fail the energy check: the force is fine, the panel is heavy, and the speed profile was set by someone who wanted the door to feel "snappy". Measure force properly. Use a gauge with peak hold, because the closing force is not constant through the travel. Hold it perpendicular to the panel at the leading edge, about half the panel height, and let the door push against it through a full closing cycle. Read the peak. Do it three times and take the highest. If you hold the gauge at an angle you are measuring a component of the force, not the force, and the number will be low and useless. Measure at the right moment, too. On most operators the door runs at full speed for the first two thirds of the travel and then drops to creep speed. The 150 N limit applies to the closing force as a whole, but the number that matters to the passenger is the one at the end, where the door meets the jamb. A peak-hold gauge will catch the highest value anywhere in the cycle, which is usually just after the door starts moving, when the motor is overcoming static friction. That peak can be 20 N higher than anything the passenger feels. Note both numbers: the peak and the value in the last 100 mm. Energy is harder to measure directly and easier to reason about. Get the panel mass from the door supplier or weigh it, get the closing speed from the controller manual or measure it with a stopwatch over a known distance, then do the sum. On retrofits with new panels and an old operator, this is the calculation nobody does. New panels are often heavier than the originals because the glass is thicker or the cladding is stainless rather than aluminum. Same motor, same speed setting, 30 percent more energy. Now the landing doors, which is where the complaints actually come from. A landing door has its own closer, usually a spring or a weight, and the car door pushes it open through the vane and roller. When the car door retracts, the landing door closes under its own closer and the car door follows. That means the closing force at a landing can be higher than at the car, because two closers are working at once, and it can be higher on the upper floors than the lower ones if the closer spring has been adjusted by three different people over fifteen years. Check a sample. On a ten-stop bank, check the bottom, the middle, and the top. If the top landing reads 180 N while the car reads 120 N, you have found a closer spring that was tightened to stop a door that was dragging, and the drag is still there. When the force is too high, the order of adjustment matters. First check for mechanical drag: sill guide, gib, rollers, hanger, and the rubber bumper at the jamb. A door that needs 40 N of extra push to overcome a dry sill guide will read high on the gauge and no parameter change will fix it. Second, check the creep speed and the deceleration ramp. Third, adjust the closing force parameter or the spring. Changing the parameter first is how you end up with a door that is within limits and still slams. When the energy is too high, you are usually dealing with mass or speed, and mass is not negotiable. Reduce the closing speed, lengthen the deceleration, and check that the door is not accelerating into the last 100 mm because a limit switch is set too late. On a heavy stainless panel, accept a slower door. A door that takes 3.5 seconds to close is not a problem. A door that takes 2.2 seconds and leaves a bruise is. One more check that costs nothing: verify the reopening device works before the panel touches anything. Stand in the doorway with your hand at knee height, and again at chest height, and let the door come at you. It should reverse without contact. On doors with a light curtain, test at the sill, at hip height, and at shoulder height, because a two-dimensional curtain mounted too high will miss a child and a trolley wheel. If the door touches you before it reverses, the sensor is misaligned or dirty and the force reading is irrelevant, because the protection is not there. Write the numbers down per landing and keep the sheet in the machine room. Next year the same argument will happen with a different inspector, and a three-year history of readings ends it in one minute.