I'm setting up a new data‑logging rig for my drag‑run sessions and need a solid, repeatable method to capture torque output throughout the launch. What sensor types and placement strategies give the most reliable readings without adding too much weight? How do you handle calibration before each run, and what sampling rates have you found balance detail and file size? Any tips on filtering noise or synchronising torque data with speed and throttle inputs would be great. Looking for a workflow that works for most setups – what’s been your go‑to approach? 🚗💨
Best Practices for Extracting Consistent Torque Data During Drag Runs
👁️ 88 görüntüleme💬 2 cevap❤️ 0 beğeni
2 Cevap
Kanka, ben drag‑run’de 6‑oslu strain‑gauge torque transdüseri şaftın hemen önüne alüminyum braketle sabitleyip kabloyu titreşimden izole ettim; kalibrasyon için her seferinde 0 Nm (fren) ve tam gazda bilinen bir tork noktasını iki‑nokta kalibre olarak kullanıyorum. Örnekleme 2 kHz yeterli, sonrasında 1‑2 kHz low‑pass FIR filtresiyle gürültüyü süzüp, aynı tetik sinyaliyle speed‑ve‑throttle verilerini aynı zaman damgasına kaydediyorum, dosya boyutu da çok patlamıyor.
I’ve been logging torque on my own drag‑run rig for the past year, and the setup that sticks out as the most reliable is a strain‑gauge‑based torque transducer mounted directly on the prop‑shaft flange. A compact, hollow‑shaft transducer (e.g., a MEMS torque sensor) keeps the weight under 200 g and gives you a linear 0‑300 Nm range with <0.5 % repeatability. I always bolt the sensor with a torque‑tightening fixture and use a short, stainless‑steel shim to preserve the original shaft geometry – that way the drivetrain sees virtually no extra inertia.
Before each run I zero the sensor with a known‑load calibration block (≈5 Nm) and run a quick “no‑load spin‑down” to capture the offset drift; storing that offset in the logger lets you subtract it automatically. I sample at 2 kHz; it’s fast enough to catch the steep torque rise during launch while keeping the CSV files under 30 MB per run. A simple 2nd‑order low‑pass filter at ~150 Hz removes the high‑frequency vibration without blurring the peak torque. For sync, I feed the engine‑ECU’s crank‑angle pulse (or a Hall‑sensor trigger) into the same DAQ channel as the torque sensor, and I log the throttle position voltage on a separate analog input – then line everything up in post‑processing by the common timestamp. This workflow has given me consistent, repeatable torque curves with minimal extra weight and manageable file sizes.