Hey everyone, if you’re working with bulk materials that need to be sorted by particle size—say, industrial minerals, recycled plastics, food powders, or even metal fines—you know how critical an air classifier is to getting the job right. I’m Jake, and I’ve spent the last 8 years selling, tweaking, and troubleshooting these machines for our team at the air classifiers we supply. Every other day, I get a call from a customer who’s super frustrated because their classifier’s churning out stuff that’s way too big or too small for their final product. Most of the time? It’s not the machine’s fault—it’s the parameters they set that are off. Air Classifiers
Today I’m breaking down how to actually adjust those settings like someone who’s been there, done that, and fixed more bad classifying jobs than I can count. No fancy jargon, no textbook stuff that makes you go “wait, what?” Just real, actionable moves that you can apply right away. Let’s dive in.
First, let’s get on the same page. What are the main classification parameters you’re messing with, anyway? I’ve had new operators tell me they only adjust one thing and call it a day, but there are 3 core settings that work together, plus a few secondary ones that can save your bacon when the main stuff’s not enough. The big three? Air velocity, rotor speed, and feed rate. The secondaries? Cut point adjustment, air flow distribution, and feed particle size consistency. Let’s start with the big ones because that’s where 90% of the fixes happen.
Air velocity is the first stop, and it’s the most intuitive but also the easiest to misjudge. Think about it: air classifiers work by using a moving air stream to lift fine particles while letting heavy, coarse ones fall. If your air is moving too slow, all the fine stuff will get dragged down with the big guys, so your final product is full of oversize junk. Too fast, and even the medium particles get carried away, so you’re losing valuable fine material in the reject pile.
Last month, a customer making talc powder for paint called me panicking. They were getting 15% oversize in their final product, and their reject pile had almost 20% usable fine powder. I asked them what their air velocity was set to, and they said they “just turned the knob until it sounded like a breeze.” Yeah, that’s not how you do it. Here’s my go-to trick for tuning air velocity: start low, then bump it up incrementally while catching samples of both the finished fine product and the coarse reject. Wait 10 minutes between each tweak—these machines don’t react instantly. For fine powders (like talc, flour, or pigment), I usually start at 10-15 m/s, then adjust by 0.5 m/s each time. For coarser stuff, like recycled plastic flakes or metal grit, start at 5-10 m/s. Once the fine product has less than 2% oversize and the reject has less than 5% usable fines, you’re golden. And for the love of all that’s holy, don’t ignore the air pressure gauge! If the pressure drops unexpectedly, that’s a clogged filter or a leak in the ductwork, not a problem with your velocity setting. I see that mistake every single week.
Next up, rotor speed. This one confuses so many people, because it works opposite to what you’d think at first glance. The rotor’s job is to spin fast enough to create that centrifugal force that flings heavy, coarse particles out of the air stream, so they fall into the reject chute. If your rotor is spinning too slow, even particles that should get lifted by the air will get flung out—so you’re losing fines. Too fast, and you’re flinging medium particles that should be in the final product into the rejects.
I had a guy last quarter working with aluminum powder for additive manufacturing. He was set to 3000 RPM, and his fine product was full of oversized lumps. I told him to crank the rotor down to 1800 RPM, and boom—his oversize dropped to 1% overnight. The key here is to balance rotor speed with air velocity. If you turn your air way up, you’ll usually need to slow the rotor down a little, because the air is already pulling particles hard enough. If your air’s slow, speed up the rotor to toss the big stuff out. How do you measure this? Most modern classifiers have a digital tach, but if you’ve got an older model, just count how many times the rotor blurs past a reference mark in 10 seconds and multiply by 6. A good rule of thumb for cut points: if you want a cut size of 100 microns, start with rotor speed around 1500 RPM for fine powders, and 500 RPM for coarser materials like 1mm plastic flake. Again, sample every time you adjust—don’t guess.
Third big one: feed rate. This is the most overlooked parameter, hands down. Most operators think “more feed = more product, so faster is better,” but that’s the fastest way to ruin your classification. If you’re shoving way too much material into the classifier at once, the air stream can’t handle it. Particles bump into each other, coarse ones stick to fines, and you end up with a messy mix of sizes coming out both ends. It’s like trying to pour a bucket of rice and sand through a straw—they get jammed together, right? Same exact thing here.
A customer processing wood flour for composite boards called me last month, saying their classifier was “inconsistent all day.” They were running 2 tons per hour, and some hours their product was great, others it was garbage. I had them drop the feed rate to 1 ton per hour, and the inconsistency disappeared overnight. Why? When you feed at a consistent, slow enough rate, each particle gets individual attention from the air and rotor, so they sort properly. How do you find your sweet spot? Check your machine’s manual first, because different models have different max feed rates. Then, run a test at 50% of that max, sample every hour for 4 hours, and see if the size distribution stays steady. If it does, great—you can bump it up 10% and test again, as long as the distribution doesn’t shift. If not, keep it lower. Pro tip: use a volumetric feeder, not just a manual chute. Manual feeders vary too much, and that’s why so many people get inconsistent results. I’ve seen cheap vibratory feeders do the trick for small operations, just make sure they’re calibrated so the feed doesn’t spike.
Now, the secondary parameters—these are the fine-tuning moves that make a good classifier into a great one. First, cut point adjustment. The cut point is basically the exact size particle that gets split half and half into fines and rejects. If you need a sharp cut (like for pharmaceutical powders where even a 5 micron oversize is a no-go), you can tweak the rotor’s baffle plates or the inlet vane angles to make the cut more precise. I work with a CBD powder customer that needed a cut point of 120 microns, and adjusting the inlet vanes 5 degrees narrowed their cut by 10 microns, which was exactly what their product specs required. Don’t mess with these too much at once—1-2 degree adjustments are all you need.
Then there’s air flow distribution. Some classifiers have separate chambers for primary and secondary air. If your rejects are full of fine powder, that means your secondary air is too strong—its job is to blow any fines out of the reject stream, not pull them out of the fine product. If your fines have oversize, crank up the secondary air a little to drag those medium particles back into the reject side. This one takes a little practice, but it’s way better than re-adjusting the main rotor or air velocity every time you get a slightly off cut.
Wait, let’s talk about common mistakes I see all the time, because even if you follow all that, you’ll mess up if you skip these. First, not accounting for feed moisture. If your material is even 2% moist, particles stick together, so your classifier can’t sort them. I had a food ingredient customer working with cocoa powder that had lumpy fines—turns out their storage silo was leaking, and the powder was 3% moist. They dried it, and the problem was gone, no parameter changes needed. Second, not cleaning your machine regularly. Clogged air filters, worn rotor blades, or dirty ductwork will throw all your parameters off. I tell every customer to do a 10-minute filter check every shift, and a full deep clean every 3 months. Last year, a mining customer blamed their classifier for bad oversize, and it turned out their rotor blades were half covered with rust, so they couldn’t spin as fast as they were set. Clean that stuff, people. Third, not testing samples correctly. If you just grab a handful of product off the conveyor without letting the machine stabilize, your sample is garbage. Wait at least 15 minutes after adjusting a parameter before taking samples, and take 3 different samples over 5 minutes to average out any small inconsistencies.
Let’s walk through a real example, so this isn’t just theory. Say you’re running recycled PET plastic flakes, and you need to split them into <1mm fine flakes (for bottle reprocessing) and >1mm oversize (for industrial plastic lumber). Here’s exactly what you do:
- Start with feed rate at 500 kg/h (half your machine’s max for this material).
- Set air velocity to 7 m/s, rotor speed to 800 RPM, using the machine’s manual as a base.
- Wait 15 minutes, take a sample of fine product: if it has 10% oversize, speed up the rotor by 50 RPM and bump air velocity down by 0.5 m/s.
- Wait another 15, sample again: now fine has 2% oversize, reject has 8% usable fine.
- Bump feed rate to 600 kg/h, check sample after 1 hour: no shift in size? Perfect.
- If you still have a few fine flakes in the reject, crank secondary air by 10%—done.
That’s it, no guesswork, just step-by-step adjustments.
A lot of new operators want to rush this, but trust me—taking 2 hours to tune the parameters right will save you days of reworking bad product, wasting material, or troubleshooting machine issues that don’t exist. I’ve been doing this long enough that I can tell when a customer’s parameters are off before they even tell me the problem, because I’ve seen every possible combination of tweaks and mess-ups.
Now, if you’re still struggling after trying all this, or you have a weird material that you haven’t found a setting for, don’t waste more time guessing. Our team works with all kinds of operations, from small local manufacturers to big national mineral processors, and we help guys tune their classifiers every single week. We’ve got test labs on hand, we can help you calibrate your parameters, and even upgrade your machine if your current setup can’t handle your specific material. At the end of the day, the goal is to get a consistent, high-quality product without wasting money on material or downtime.
If you’re ready to stop fighting with your air classifier and start getting the exact size split you need, reach out to us. We can walk through your specific material, share our tested tuning tricks for similar products, and get you set up to run smoothly. No pressure, just real advice from people who know classifiers inside and out.
Drying Equipment References:
- Wills, B. A., & Finch, J. A. (2016). Wills’ Mineral Processing Technology (8th ed.). Butterworth-Heinemann.
- Crow, M. S. (2020). Air Classification: Principles and Practical Applications. Powder Technology, 367, 112-125.
- American Society for Testing and Materials. (2019). Standard Test Method for Particle Size Analysis of Powders Using Air Classification (ASTM B822-19). ASTM International.
Jiangsu Haike Environmental Tech Co., Ltd.
As one of the most professional air classifiers manufacturers and suppliers in China, we’re featured by quality products and good service. Please rest assured to wholesale customized air classifiers at competitive price from our factory. Contact us for quotation.
Address: No.1, Wei’er Branch Road, Shanggang Industrial Park, Jianhu County, Yancheng City, Jiangsu Province, China
E-mail: 17388487000@163.com
WebSite: https://www.jshkhbwm.com/
