Wind turbines

[gar]

130425-2150 EDT

This evening our local SME (Society of manufacturing Engineers) chapter visited the Dowding Industries plant in Eaton Rapids, Michigan.

Here they have what is probably the longest bridge mill in the world. The base is 106' long and they can mill a part up to 96 ft long. To date the longest part has been 65'. The primary purpose of the machine is to machine the form for making wind turbine blades. The accuracy is on the order of 0.001" over the 96'. It is a 5 axis MAG machine.

Another special CNC machine they have has 3 heads, each with a 100 HP spindle motor, and a central rotary table. This machine is use to machine the raw casting for the turbine rotor hub. The raw casting is about 45,000 #, and after machining about 40,000 #. A single hub can be machined in about 6 hours where previously a single spindle machine required about 26 hours. The machine cost $14,000,000.

Another machine, not at this location, is a CNC to lay up the carbon fiber sheets on the blade form for turbine blades.

These machines are helping to reduce the cost of a wind turbine. and therefore the cost to generate electricity.

Currently the hub castings come from China or Finland, and besides transportation costs there is about $14,000 spent just to package the casting for shipment. Plans are in the works to build a casting plant just across the street from the machining plant. This will be the first new casting plant in the US in about 40 years. Big cost savings here.

There are a considerable number of wind farms being installed in Michigan. So some of these parts are directly used in Michigan.

Since the thruput of the machines is much greater than demand the machines get used to make other large objects. There is continuous demand for large parts used in fracking. This is continuous because the parts only last about 300 hours.

.

 

[eHunter]

Sounds like good news for local employment growth and the US wind industry and nirvana for a machine tool junkie(me).
The bridge mill must be massive to have the rigidity to maintain the 0.001 in. accuracy.

 

[gar]

130426-0845 EDT

eHunter:

The center photo at http://www.dowdingmachining.com/Services.aspx is of the long bridge mill. It is impressive to see a machine table that is 106 ft long and about 8 ft wide that is flat within about 0.001". There is a large rack gear on each side of the table. Also a chip conveyor on each side. I was not not able to determine what kind of position encoder is used for the x-axis. The operator and controls ride with the bridge assembly.

.

 

[K8MHZ]

Here are some pics of parts of some of the largest wind turbines in the world passing through my town.

www.photobucket.com/turbineblades

The trucks in the pics are over 200 feet long.

They came into our port in Muskegon, on the other side of the state from Gar, and were taken to a new wind farm in the middle of the state. The blades came from Germany. The hubs and generators (nacelles) are made here in the US.

We are supposed to be getting a fairly good sized wind farm here in Muskegon, about 3 miles from my house. I think we are just one study from being fully approved.

 

[rsnell22]

I think wind power is a cruel joke. A wealthy investor's tax shelter' and very, very little else.

Just try to find a quoted price per KWH.

Everyone I've tried to ask, from power companies to people who work on the junk, just end up yellling at me.

Anything that has to rely on legislation to get utilities to buy into it, can't be reliable, or economically viable.
If it was so good they would put the stuff up without legislation or government subsidies.

 

[cmreschke]

Agree. But still happy to work on the turbines in the thumb in Michigan. Hope to get out on the next site this fall.

 

[K8MHZ]

I think wind power is a cruel joke. A wealthy investor's tax shelter' and very, very little else.

Just try to find a quoted price per KWH.

Everyone I've tried to ask, from power companies to people who work on the junk, just end up yellling at me.

Anything that has to rely on legislation to get utilities to buy into it, can't be reliable, or economically viable.
If it was so good they would put the stuff up without legislation or government subsidies.

I went to one of the meetings the county had with the makers of the turbines we are supposed to get. They said they could make it for 8.2 cents and make a comfortable profit at 12 cents.

The company is not selling the electricity to our county. It is leasing the land (for big bucks) from the county to build the farm from which they will sell the electricity to the POCO.

 

[gar]

130426-2329 EDT

Dowding in discussions with their customers say the customers see $0.06/kWh as a feasible value for turbines being built from the latest production and technical developments. Dowding has created blades with greater energy conversion efficiency, and have reduced the cost of making the parts they make. So their contribution helps reduce the cost. Also turbines are being put higher in the sky, more wind there.

Current cost to generate electricity from existing coal fired power plants is about $0.03/kWh. But these plants are being forced out of existence by government rules and regulations. To some extent it is virtually impossible to build new coal fired plants in the US.

The "green" people think you can get all your electrical energy from renewable sources. That is not realistic. Some of our electrical energy can come from renewable sources, but not all. And probably not most. So where is it going to come from. Nuclear is probably the best source at this time. But do it like France and recycle the waste.

Note: that the high percentage of renewal by some northern European countries is possible by backup from France's nuclear capability. When those countries can't produce, then they get energy from France.

.

 

[mike_kilroy]

these servos use simple incremental 1Vpp sine encoders by Heidenhain inside the Siemens motors. There is no limit on the length of the rails for the U5 machine or the other table-in-center bridge profilers; typical lengths are 200-300 feet for making Boeing and Airbus wings. Same accuracy and repeatability. There are two motors on the X axis that move it down the rack/pinions on these machines, each controlled independently by a CNC axis in the control to not skew by more than about .006" max side to side. When higher speed is required, the bridge type gantry's name changes and linear motors are used instead for like 2000IPM max traverse speeds. There are 1-5 large HP spindles per machine across that moving gantry, depending on the machine. I love working on these and supplying parts for them here in Cincinnati, OH!

 

[gar]

130428-1126 EDT

mike_kilroy:

That would mean that fundamentally the racks and pinions are the fundamental encoders and their accuracy and resolution will determine the system accuracy and resolution.

Can you tell us more about the accuracy and performance of these machines?

.

 

[mike_kilroy]

From the 1970's to early 2000 we (& others) supplied single servo motor with resolver feedback (1 rev of resolver = 0.200 inches travel) for these X & X' axes... the dual pinions (about 15" apart) were fed via a very precision and complex antibacklash gearbox (about 2'x3') on each side of the machine. The workmanship and precision parts used allowed aligning each rack piece such that they would get the overall accuracy and repeatability down in these small numbers.

In the early 2000's (IIRC) the $50,000.00 gearbox was replaced with our precision low backlash planetary gearheads - on TWO motors per side. These planetary units have flat output surface that the pinion is mounted directy on (so no shaft). Now the CNC has FOUR servo axes for X: 2 for each side, and the CNC not only sends the same basic command to each motor, but build in an antibacklash constant current to wind up both pinions on a rack to remove the high r/p backlash. So still no backlash to contend with, helping the position accuracy. In the older systems the CNC sent +/-10v velocity command to our servodrives and gain was adjusted with Potentiometers. If you are familiar with mach tool gains, you will recognize when I say these ran at a gain of 1 (in/min/mil). Nowadays the typical Siemens 840D CNC takes care of velocity loop too I think, the analog command isreplaced with digital sigs, the 200IPM speeds that got to 400IPM are upto much higher, and pos loop gain is probably 3-4 times higher. All the CNCs get laser shot and comp tables written from this data to correct for repeating/cyclic position errors based on the mechanics. The encoders are typically 1024PPR and these sine/cosines are divided way down for much higher resolution; some of our drives divide by up to 4096. We always shot for at least 10-20 counts or more per repeatability required. There are occassions where absolute encoders are used not I do not believe the U5 used them. If absolute, it is commonly also Heidenhain proprietary Endat, which is simply replacement of Z (marker) channel output with data line that can be asked for 12bits of absolute pos data on command, which the control then just adds to the other 12bits of incremental 1Vpp encoder data - like at power up; otherwise, it is still just sine encoder technology for everyday positioning. Due the long travels, linear feedback (encoders, scales, lasers) is not used on a lot of these large machines. all this just from memory so hope i am not wrong too badly on details! Lastly, Mag just sold to a french engineering company (five - sp?) their machine tool manufacturing company.

here are some links to more details on these neat machines:

http://www.mag-ias.com/en/mag/products-services/milling/gantry-and-portal-mills/u5-portal-mill.html

[URL="http://www.mag-ias.com/en/mag-news/press-archive/archive/archive2/press-archive/2010/dual-spindle-5-axis-gantry.html"]http://www.mag-ias.com/en/mag-news/press-archive/archive/archive2/press-archive/2010/dual-spindle-5-axis-gantry.html

[URL="http://www.mag-ias.com/en/mag-news/press-archive/archive/archive2/press-archive/2012/article/article/aerospace-dynamics-international-launches-major-capacity-increase-in-titanium-machining-with-order-f.html"]http://www.mag-ias.com/en/mag-news/press-archive/archive/archive2/press-archive/2012/article/article/aerospace-dynamics-international-launches-major-capacity-increase-in-titanium-machining-with-order-f.html


[/URL][/URL]