Mini-Split Flare Torque and Flare Size Field Card

Why this matters

Most mini-split leaks a shop goes back for are at a flare, and most of those flares were tightened by feel. An under-torqued flare seeps slowly and shows up as a low-charge call weeks later. An over-torqued one cracks the flare at its root or splits the nut, sometimes during the first heat cycle when the joint expands. Both look fine the day of install and both pass a quick bubble test. The torque wrench is the difference, and it only helps if the number set on it is the one the maker printed for that nut on that tube.

This card owns the torque and flare-size numbers and why they are what they are. The full procedure for cutting, reaming, flaring and joining is in How to Make and Torque a Flare on a Mini-Split Line Set.

The table as makers print it

Every mini-split installation manual carries a flare table. The values below are the ranges commonly printed for the four tube sizes residential systems use. The maker's table for the unit on the truck governs; where it differs from this card, use the manual and note the manual's value on the ticket.

FLARE TABLE (typical, as printed in install manuals)

TUBE OD   TORQUE N-m   TORQUE ft-lb   FLARE A dim, mm
-------   ----------   ------------   ---------------
1/4 in    14 to 18     10.3 to 13.3   about 9.1
3/8 in    34 to 42     25.1 to 31.0   about 13.2
1/2 in    49 to 61     36.1 to 45.0   about 16.6
5/8 in    68 to 82     50.2 to 60.5   about 19.7

ft-lb = N-m x 0.7376
A dim = flare outer diameter at the face, 45 degree flare

The ft-lb column is the N-m column multiplied by 0.7376 and rounded to a tenth; check it against your wrench's scale rather than memorizing either. The A dimensions are the values commonly printed for R-410A and later refrigerants, with a small minus tolerance; some manuals print slightly smaller figures for older refrigerants and older tools. Your manual's figure and tolerance govern.

Why the torque climbs faster than the tube size. The nut has to compress a larger flare face against a larger cone, and the larger nut has a larger thread diameter, so the same clamping pressure on the sealing face takes more twisting force. Doubling the tube size from 1/4 to 1/2 inch roughly triples to quadruples the torque, which is why "about the same feel, a bit harder" undertightens the big nuts and overtightens the small ones.

Why a range and not one number. The range covers the spread in nut and tube tolerances across the maker's own parts. Aim for the middle of the printed range unless the manual says otherwise. The bottom of the range is not a target you can reach by feel and be sure of.

Why a torque wrench and not feel

A flare seals because the soft copper face is pressed against the hard brass cone of the fitting until the surfaces conform. Too little force and the faces never fully conform; refrigerant finds the gap at operating pressure, more so on the high side in cooling and on the vapor line in heating, where pressure and temperature swing hardest. Too much force thins the flare at its root, where the tube bends 45 degrees into the flare, and work-hardens it. That root crack is the failure that shows up after a few heat-and-cool cycles rather than on day one.

Field experience with the same nut varies by a factor of two between techs. A 1/4 inch nut is small enough that a strong tech with a long open-end wrench passes the top of its range without feeling it. A 5/8 inch nut is big enough that a tech used to 3/8 work stops well short. The only way to get the same result from every installer is the number on the wrench.

Backup wrench every time. Hold the valve body or union with a second wrench while you torque. Without it, the torque goes into the service valve's mounting or twists the indoor unit's stub tube, and the reading on your wrench no longer says what reached the flare.

Flare nut types

Standard short flare nuts are what most units ship with on their service valves and indoor stubs. Use the nut the maker supplied with the unit; it matches the valve and the flare dimension the table assumes.

Wider-hex nuts for higher-pressure refrigerants. Nuts supplied with R-410A and later equipment usually have heavier walls and a larger hex on the bigger sizes than nuts from older refrigerant equipment. Do not reuse an old nut from a removed system on a new one; the hex and thread may fit the valve while the nut's wall and internal cone are made for a different pressure and flare.

Long or forged nuts are offered as replacements and carry the thread farther back along the tube, which gives some support behind the flare where vibration fatigue starts. They are fine where the maker allows them, and they must still meet the maker's flare dimension.

A nut that was over-torqued is scrap. Do not re-use it with a fresh flare. The thread and internal cone have been stretched, and the next flare will not seat the same.

Worked example: an 18,000 Btu/h single-zone install

An 18,000 Btu/h ductless heat pump with a 1/4 inch liquid line and a 1/2 inch vapor line, 25 feet of line set, joined at the outdoor service valves and the indoor unit's stubs. That is four flare joints: two at 1/4 inch, two at 1/2 inch. The manual prints:

• 1/4 inch: 14 to 18 N-m, so the middle is 16 N-m.
• 1/2 inch: 49 to 61 N-m, so the middle is 55 N-m.

The shop's click-type torque wrench reads in ft-lb.

Convert. 16 N-m x 0.7376 = 11.8 ft-lb. 55 N-m x 0.7376 = 40.6 ft-lb.

The crowfoot correction. The outdoor valves are tight against the cabinet, so the tech uses a crowfoot (an open-end head that fits on the wrench's square drive) to reach the nut. A crowfoot mounted in line with the handle adds its own length to the lever and the nut sees more torque than the wrench reads. The correction for an in-line crowfoot is:

setting = target x L / (L + E)

L = wrench length, drive center to handle grip center
E = crowfoot length, drive center to nut center

1/2 in flare:  40.6 x 12 / (12 + 1.5)
             = 40.6 x 12 / 13.5
             = 36.1 ft-lb setting

1/4 in flare:  11.8 x 12 / (12 + 1.0)
             = 11.8 x 12 / 13.0
             = 10.9 ft-lb setting

L here is the shop wrench's 12 inches; the 1/2 inch crowfoot measures 1.5 inches drive to nut center and the 1/4 inch measures 1.0 inch. Measure your own; these are this kit's dimensions.

Turn the crowfoot instead. Set the crowfoot at 90 degrees to the handle and the effective length does not change, so no correction is needed. Where the space allows it, that is the cleaner way. The tech in this example rotated the 1/4 inch crowfoot to 90 degrees and set the wrench to 11.8 ft-lb directly; the 1/2 inch crowfoot would not clear the cabinet at 90 degrees and went on in line at the corrected 36.1 ft-lb.

What a skipped correction would have done. At the 1/2 inch joint, setting 40.6 with the crowfoot in line delivers 40.6 x 13.5 / 12 = 45.7 ft-lb, which is 62 N-m (45.7 divided by 0.7376). That is just past the top of the 49 to 61 N-m range. One crowfoot, ignored, moved a mid-range joint out of the range.

Then the flare dimension. Before each joint, the tech checked the flare's outer diameter against the manual's A dimension with a flare gauge. The first 1/2 inch flare read small, the tube had not been pushed far enough into the flaring bar; it was cut off and re-made. A short flare seated at full torque leaves less sealing face and more stress at the root.

What changes the answer

The manual prints a different number. Some makers print a single value with a tolerance instead of a range, and a few print different torques for the indoor and outdoor ends. Use theirs.

Oil on the flare. Makers differ on whether to put a drop of refrigerant oil on the back of the flare (where the nut bears) and on the threads. Oil reduces friction, so the same wrench setting delivers more clamping. Where the manual calls for oil, use the oil type it names and only where it says; never on the sealing face unless directed. Where it says no oil, use none.

An A2L system. Field-made flares in the occupied space on A2L equipment are subject to the maker's instructions and to the safety standard and code your jurisdiction adopted for that refrigerant; some installations require a joint type rated for the purpose indoors. Check before you flare indoors. See related: A2L Equipment Install Requirements Field Card.

Re-making a joint that was opened. A flare that has been torqued and then loosened has already been formed against one cone. Most makers say to cut it off and re-flare rather than re-torque the old one.

How to verify each flare

• Gauge the flare against the A dimension before assembly: round, centered, no splits, no burr on the face.
• Wrench set to the table value, corrected for any in-line crowfoot, with a backup wrench on the body.
• Mark the nut. A paint or marker line across nut and body after torquing shows later whether the nut has moved and tells the next tech the joint was torqued.
• Pressure and leak test the joints under a standing nitrogen test before evacuation (see How to Run a Standing Nitrogen Pressure Test on a New Line Set), never above the test pressure on the unit's manual.
• Keep the wrench in calibration. A click wrench stored set at a high value loses accuracy; back it to its lowest setting when it goes back in the case.

References

• Mini-split installation manual for the unit installed: flare torque table, flare A dimension, oil instructions.
• See related: How to Make and Torque a Flare on a Mini-Split Line Set; How to Run a Standing Nitrogen Pressure Test on a New Line Set.
• See related: A2L Equipment Install Requirements Field Card; A2L Refrigerant Handling (R-454B and R-32).
• Torque wrench maker's instructions for adapter length correction.