Aluminum needs a coarser blade than steel at the same thickness, because the failure mode is not tooth wear, it is a packed gullet. This selector takes your material, section and cut length and returns the tooth pitch, band speed and feed rate that keep chips clearing, along with the teeth-in-cut check that tells you whether the blade will load up before it dulls.
Last updated July 2026
Many horizontal production bandsaws top out near 400 SFM. Set the machine maximum to what your saw actually does, not what the alloy wants, and the recommendation will be clamped to it.
Base values for a bi-metal M42 blade with flood coolant on a machine that can reach the speed. Multiply by the blade and coolant factors in the table below.
| Alloy and temper | Band speed (SFM) | Cut rate (sq in/min) | Notes |
|---|---|---|---|
| 6061-T651 | 4,500 | 35 | Baseline. Chips break well, tolerant of feed error. |
| 6063-T5 / T6 | 3,500 | 30 | The worst stringer. Coarsest pitch you can get away with. |
| 5052-H32 | 4,000 | 28 | Gummy. Long chips need gullet room. |
| 5083-H321 | 4,000 | 28 | Gummy. Go one pitch coarser than the chart says. |
| 5000 series cast tool and jig plate | 5,000 | 40 | Fastest of the group. Cast structure breaks chips cleanly. |
| 2024-T351 | 4,000 | 25 | Copper bearing, moderately abrasive. |
| 7075-T7351 | 3,800 | 24 | Overaged, slightly kinder to the tooth than T651. |
| 7075-T651 | 3,500 | 22 | Harder and more abrasive. Slow the band, keep the feed up. |
| 7050-T7451 | 3,500 | 22 | Thick-section aerospace plate. Treat like 7075-T651. |
Section length in the cut, S, is the instantaneous depth of material the blade passes through, not the length of the workpiece. Every pitch below is one step coarser than the standard steel recommendation for the same section.
| Section in the cut S (in) | Variable pitch | Constant pitch alternative | Teeth in cut at midrange |
|---|---|---|---|
| Under 0.25 | 10-14 | 10 TPI | 2.4 at 0.20 |
| 0.25 to 0.50 | 8-12 | 10 TPI | 3.8 at 0.375 |
| 0.50 to 1.00 | 6-10 | 8 TPI | 6.0 at 0.75 |
| 1.00 to 2.00 | 4-6 | 4 TPI | 7.5 at 1.50 |
| 2.00 to 3.00 | 3-4 | 3 TPI | 8.8 at 2.50 |
| 3.00 to 5.00 | 2-3 | 2 TPI | 10.0 at 4.00 |
| 5.00 to 8.00 | 1.4-2 | 2 TPI | 11.1 at 6.50 |
| Over 8.00 | 0.75-1.25 | 1 TPI (special order) | 10.0 at 10.00 |
| Blade width | Kerf (in) | Squareness (in per in) | Typical use |
|---|---|---|---|
| 1/2 in | 0.035 | 0.020 to 0.030 | Small horizontal saws, tight radius work on a vertical saw. |
| 3/4 in | 0.042 | 0.015 to 0.025 | General shop horizontal saw up to about 6 in section. |
| 1 in | 0.050 | 0.012 to 0.020 | Production horizontal saw, plate and heavy bar. |
| 1-1/4 in | 0.063 | 0.010 to 0.018 | Large production saw, thick plate on edge. |
| 1-1/2 in | 0.083 | 0.010 to 0.015 | Heavy plate saw. Widest kerf, straightest cut. |
| Factor | Band speed | Cut rate | Notes |
|---|---|---|---|
| Bi-metal M42 blade | 1.0x | 1.0x | The default. Best cost per cut for aluminum. |
| Carbide tipped blade | 1.8x | 1.6x | Worth it on 7075 and 7050 in volume. |
| Carbon steel blade | 0.5x | 0.5x | A hobby-saw answer, not a production one. |
| Flood coolant | no change | 1.0x | Baseline. Flush the gullet, not just the blade. |
| Mist coolant | no change | 0.8x | Acceptable compromise. Aim at the tooth entry side. |
| Dry | no change | 0.5x | Aluminum welds to the tooth face. Expect to lose the blade. |
S is the instantaneous depth of material the blade passes through. Plate laid flat: S equals the cut length, because the full width stays engaged for the whole cut, which is exactly why that is the bad orientation. Plate stood on edge: S equals the thickness. Solid round: S equals 0.7 times the diameter as an average through the cut, with the peak at the centerline equal to the full diameter. Rectangular bar: S equals the smaller of the two cross-section dimensions if it can be oriented that way, else the larger. Tube: S equals 2 times the wall thickness. Angle: S equals the leg thickness.
TIC = S x TPI_average, where TPI_average is the mean of the variable pitch range, so a 4-6 blade has a TPI_average of 5. The target band for aluminum is 3 to 6. Acceptable is 3 to 12. Above 12 the gullets pack, below 3 the teeth can straddle the section and strip.
Choose the variable pitch designation whose TPI_average puts the teeth in the cut nearest 4.5, subject to the teeth in the cut never dropping below 3. The section chart above is the conventional lookup and will sometimes land one step finer, which is still workable, just closer to the packing limit.
V = base_SFM(alloy) x blade_type_multiplier, then clamped to the machine maximum. If the uncapped value is above your machine maximum, the tool reports the clamped speed and says so, because there is no point specifying a speed the saw cannot reach.
R = base_rate(alloy) x blade_type_multiplier x coolant_multiplier, in square inches per minute. Cut area for a plate cut is A = cut_length x thickness. For a round, A = pi x d squared / 4. For a tube, A = pi x (OD squared minus ID squared) / 4. Cut time in minutes is t = A / R, reported in seconds when it falls under 120.
f = R / S, in inches per minute, where S is the section length in the cut. This is the number to dial into a saw with a feed rate display, and it is why the same cut rate gives a very different downfeed on a plate laid flat than on the same plate stood on edge.
total_loss = kerf_width x number_of_cuts, with the kerf width read from the blade width table. On a 1-1/2 inch blade at 0.083 inch kerf, twenty cuts throw away 1.66 inches of material before you account for facing.
Anything tighter than plus 0.030 inch on length or 0.005 inch per inch on squareness is a mill operation rather than a saw operation. Need the stock to cut these parts from, get an instant quote on 6061, 7075 and 7050 plate cut to your size, with the kerf and the saw tolerance already accounted for.
Go coarser than you would for steel at the same thickness. For 1 to 2 inch aluminum use a 4-6 variable pitch, for 2 to 3 inch use 3-4, and for anything over 3 inch use 2-3 or coarser. The target is 3 to 6 teeth in the cut, against 6 to 12 for steel, because aluminum fails by packing the gullets rather than by wearing the teeth. Anything finer than about 6 TPI on thick aluminum will load up and start cutting crooked.
Around 4,500 SFM for 6061 with a bi-metal M42 blade, and about 3,500 SFM for 7075 and 7050 because they are harder and more abrasive. Carbide tipped blades run roughly 1.8 times those speeds. Many horizontal bandsaws top out near 400 SFM, so check your machine before chasing these numbers, and if you are capped low, drop the feed proportionally rather than forcing the cut.
Almost always too fine a pitch, too little lubrication, or both. If the gullets fill before the tooth exits the cut, the chip has nowhere to go, it packs, heat builds, and aluminum welds to the tooth face. Count your teeth in the cut: multiply the section depth by the average TPI, and if the answer is over about 12 you need a coarser blade. Then check that you have flood or at least mist coolant, because aluminum cut dry will weld to the tooth every time.
On edge, whenever the machine allows it. A 12 inch wide 1 inch plate laid flat presents 12 inches of section to the blade for the entire cut, which puts roughly 60 teeth in the cut at 4-6 TPI and packs every gullet. Stood on edge the same plate presents 1 inch, which is about 5 teeth in the cut and exactly where you want to be. Same blade, same plate, and the difference is whether the cut works.
Realistically minus 0.000 plus 0.060 inch on length for sections under 2 inches, and minus 0.000 plus 0.125 inch on larger sections, with squareness of 0.010 to 0.030 inch per inch of section. Surface finish runs 250 to 500 Ra. If your print needs tighter than plus 0.030 inch on length or 0.005 inch per inch of squareness, you are ordering a milling operation, and it should be specified and priced that way rather than argued about after delivery.
You can, but you will cut your blade life roughly in half and the finish will suffer. Aluminum is soft enough not to wear the tooth, but it welds to the tooth face and builds up until the tooth stops cutting and starts rubbing. Flood coolant is best, mist is an acceptable compromise at about 80 percent of the flood cut rate, and dry cutting drops you to roughly half rate with a real risk of the blade seizing in the cut.