HAWT on crankup/tiltover tower

[Stewart Corman]

First glimpses of completed assembly of the 3 blade/single rotor version of my tiltup furl head 10 foot diameter HAWT mounted on 62' guyed tower:

551

the rest of these photos are posted here:
http://picasaweb.google.com/stewcorman/10TurbineMounted#


I have several 15 second video clips posted on YouTube as well:

YouTube - Wind Turbine 10ft

head spinning closeup:
YouTube - wind turbine

crankup section lowered:
YouTube - HAWT crankup/tiltover tower - cranked down

slight tilt to balance point:
YouTube - HAWT crankup/tiltover tower - slight angle

tilted horizontal:
YouTube - HAWT crankup/tiltover tower - horizontal

full tiltover:
YouTube - HAWT crankup/tiltover tower

It is currently wired to the house 3 phase, rectified to DC, inline shunt resistor to measure current (requires voltage amplification), 4 channel Dataq A/D to record voltages (using 40:1 resistor bridge for 0->20V range) at 15 second averaged intervals directly into laptop.

A rotary switch to enable varied loading:
shorted, open NL, 100ohm, 200ohm,300ohmm, 400ohm nichrome wire resistance loads.

Servo motor is driven by 8.4:1 sproket chain drive, rated at 460v/1.5KW AC at 3000rpm.
I have measured 400v+ DC NL already and seen the head lift up smoothly.

Next project is to mount/wire LaCrosse weatherstation on separate 60 foot TV antenna tower to collect WS data in one minute averaged periods.

Stew Corman from sunny Endicott

[Rob Beckers]

Stew, what an absolutely magnificent tower! :first:
Congratulations on the construction (and its completion, I know you've been working on it for a while). It's no small feat to build a 62' tilt-up that can be lowered and raised with such ease.

-RoB-

[Stewart Corman]

Rob,
Mario inspired me, but his is balanced better and easier to crank up ...all are easily to crank down :D

guess another 200# of lead wouldn't hurt

Stew

[Gil Martino]

Wow, nice job Stu.

Stu introduced me to this web site and I am intrigued by so many interesting and power producing ideas. I love the wind power!.
Gil

[Nicolas Fournier]

congratulation for the completion of your project !
Its been a while that im reading on this board and i've followed your achivements.


is the use of a tower like yours was more costly than using a tilting tubing tower, or where you limited by the space/foot print ?


great job !

[Stewart Corman]

The electrical connnections to the house are completed inside.
The generator is a PM 3 phase servo motor rated at 460v AC at 3000 rpm and 1.5KW.
There is an 8.4:1 roller chain sprocket drive, so that at turbine rotor of 360rpm, the DC output will be 650v.

There are 4 nichrome elements housed in a SS rack assembly and they are hardwired in series. A heavy duty rotary switch picks off the connection points to allow a varied load for a particular set of readings.

Overall view of all the connections are as follows:

554

The wild 3 phase AC comes into a buss bar, then runs through a full wave bridge:

555

A rotary switch enables NL, 100, 200, 300, 400 ohm nichrome resistance heating elements, or shorted (lockup brake) and a precision shunt resistor ( requires instrumentation amp) allows current measurement:

556

A 115v AC power relay trips the load on after the turbine spins to about 35rpm NL and is triggered off a 40:1 resistor divider into a SS relay. I am thinking of adding a diode and cap to add hysterisis and restrict "bounce". Ultimate operation would include a variable load provided by a PWM of a power darlington pair, so turbine starts under no load, and gradually increases load up to a design point determined emperically from data collected with the four load options tested here.

557


Data is collected by a Dataq A/D which is run in DC mode with a range of 0->20v DC with 10 bit resolution. One is being reserved for current logging.
Only one of the four channels is currently hooked up to monitor DC voltage output and a voltmeter simultaneously monitors the live raw output for any one resistance load chosen. The voltage divider is calibrated for a known input voltage:

558

The laptop can record up to four live variables and will be set at averaging 15 second intervals.
Since voltage is linear with rpm, every 100 volts = 55rpm, and no need to monitor frequency. A scope is available to view AC waveforms.
The weatherstation will record at one minute intervals and then 1750 readings can be downloaded from the buffer along with timestamps.

I still cannot conceive of how to sensor data and will be taking suggestions from others on the board. Out of 1440 readings per day when the wind is blowing, there are instances where turbulance and change of direction will cause the tail to pivot and readings will NOT be indicative of the recorded WS. I will suppose that to take the maximum output at any given WS during short intervals ( 5 min??) will be the first approach.

Please chime in if you have any good thoughts on this.

Stew Corman from sunny Endicott

[Rob Beckers]

Good stuff Stew!
Where did you get the nichrome resistors for the diversion load? Those look pretty substantial.

For logging wind speed vs. output there is a standard (actually there are a few). Take a look around the NREL site, there must be a doc on it somewhere. Going from memory, so I'm probably wrong; the idea is to log output and wind speed every few seconds, then average that data into 'bins' of a number of minutes. If memory serves me the two methods are to use 2 minute or 10 minute bins. Those data points are then used to map a curve through.

-RoB-

[Stewart Corman]

Rob,

Where did you get the nichrome resistors for the diversion load? Those look pretty substantial.
those came 25+ years ago from the salvage yard taken out of a SS commercial oven/test chamber that was scrapped out...However, I was just looking at one the other day and will be shopping there for something else later this week... maybe I should stockpile the elements since they usually are segmented ie mine are two 50 ohm coils per rack which I hardwired in series, but in parallel it would be 25 ohms and two racks in parallel would be 12.5ohms ( we are talking multiple KW ratings here) ..they can be very versatile

Back on topic:

the idea is to log output and wind speed every few seconds, then average that data into 'bins' of a number of minutes.my concern about data sensoring is not in "binning" , but an algorithm that can sort thru 1000's of data points and remove statistical "outliers" ...can't get any more WS data that what the La Crosse delivers ...my logic would be that you can't get out any more power than the maximum you see at any WS, the minimums have too many bogus artifacts ....I like controlled experiments and that requires a full size wind tunnel !!

Stew