This is great info. Thanks. One of the things I really appreciate about this group is the willingness of everyone to share their knowledge and experience with those of us who are not so knowledgeable and experienced.
Steve, should I see one or three injection pulses per revolution? Is this trace for 5000 or 1666 RPM?
Steve’s data suggests that I have been working under an incorrect understanding of the injection process. Is fuel injected:
1. During a single pulse (per combustion event) with pulse duration determining fuel delivery, or
2. During multiple (fixed duration) pulses per combustion event with the number of pulses determining fuel quantity delivered, or
3. Some combination of 1 and 2?
If I read Steve’s data correctly, No 1 is the case and a fuel totalizer/ flow rate instrument could be based on a look-up table or a functional relationship between pulse duration and total flow (perhaps adjusted by manifold pressure). Steve, how did you determine injector “dead time”? Did you measure it directly or use statistical methods? Do you find this value to be constant, independent of engine operating parameters, for a given injector or is it variable? If variable, what other parameters do you consider?
Once again, thanks.
Gordon C. Alling, Jr., PE
President
acumen Engineering/Analysis, Inc.
540-786-2200
www.acumen-ea.com
From: Rotary motors in aircraft [mailto:flyrotary@lancaironline.net] On Behalf Of Steven W. Boese
Sent: Sunday, April 21, 2013 2:39 AM
To: Rotary motors in aircraft
Subject: [FlyRotary] Re: Fuel injector pulse frequency
Gordon,
The attached image shows signals generated by an EC2 for a 13B engine at 5000 RPM. The signals are identified by color in the legend at the right of the image.
Like Ed, I have built fuel totalizing and flow rate instruments based on discrete counter IC's and based on a microprocessor. The unit based on discrete counters was accurate only under the conditions at which the calibration was made. The reason for this is the effect of the dead time of the injector preventing a simple correlation of the length of an injector pulse with the amount of fuel delivered during that pulse when the length of the pulse is variable. Since the aviation application often involved operation at the conditions used for calibration, this limitation was not often very significant.
The system using the microprocessor enabled the accurate calculation of the amount of fuel delivered by each pulse at any injector pulse length by taking into account the injector dead time. The injector dead time was determined for the injectors used and its value was part of the programming. The microprocessor version uses inputs from the primary and secondary injectors and displays the RPM, fuel flow rate, and the amount of fuel remaining in each tank since I have a duplex fuel valve with a switch telling the microprocessor which tank is in use. I chose to use the Texas Instruments Launchpad development system which cost about $5 and included two microprocessors. Of course, other components were required for the final unit, the most expensive part being the display.
RV6A, 1986 13B NA, RD1A, EC2