Personal CNC?

There has been a thread going through my mind involving the general field of machinery.   The design of specialty machinery requires a great many disciplines, truly a mechatronic endeavor.

Over the years, machine tool makers constantly worked on making the machines more complex in order to serve the market with greater functionality.   In fact the goal seemed to be to make the machines and control systems more complex so that one machine could solve a wider range of geometry problems.  Unfortunately, this leads to ever increasing cost.  Take a simple three axis mill and add a fourth or fifth axis to it and it’s not just the cost of the additional axes of motion that will impact the final cost of the machine.  It’s the complex mechanics needed to support the fourth and fifth axis and articulate their geometry correctly PLUS the two extra servo motors and their respective feedbacks AND a huge programming effort to make sure that the coordination of the axes is as precise as expected.

And the mechanics have to be as accurate and reliable over hundreds of thousands of operations so that the specified precision of the machine does not deteriorate over time.  So things start getting pretty complex.  And when you make a machine tool that is going to cost $100,000 or more, you can’t afford the problems of a design that won’t hold up in production.  So you do a lot of testing to verify performance, which usually involves a lot of custom measurement equipment and a lot of manpower and development time.

But what if you reverse the goal of the design process? What if the objective were to create a machine tool that has the lowest cost for a specific set of features.  Let’s face it, if you know that you will not need 10’s of thousands of parts per year, or if the precision tolerances are not extremely tight, you can get a lot done on a budget.

Machinery cost is only the beginning of the equation.  Amortization of the cost of the machine over the number of parts to be produced is critical to holding cost down and making a profit.  That’s where the paradigm shift creates value.  Lower cost also means smaller batch size when calculating break even points.

So the discussion of how to make a cheaper machine tool must be considered in it’s proper context.  And history proves that it works because that’s what the folks at Haas did some years ago.  They came up with high quality machine tools that cost $50K, roughly 1/3 the cost of the available technology.  This opened up a whole new playing field in the CNC industry.  They did the job so well, that they now do business all over the world with one of the most cost effective pieces of equipment around.

And now for the next wave.  Tormach is producing a high quality 3 axis machine tool at a $10,000 starting price. Full CNC control.  And there are others available from China and India, which while not to be compared on precision, may be exactly what a small company needs to get their product to the market cost effectively.

So the real trend is just getting started and will give rise to whole new layers of improved cost and performance.  Personal fabrication technology is emerging all over the US through innovative small companies who are solving the most important problem of all.  Bringing new products to the market cost effectively.  I think there’s going to be some great opportunities.

20% Wind Power?

The DOE has published a 248 page document “20% by 2030″ providing a technical background of wind power and a roadmap for creating 20% wind energy supply in the US electrical energy mix.  At last week’s Renewable Energy conference there was an update from three of the consulting firms that have been providing input for the Department of Energy.  The firms are highly qualified engineering and technical firms with excellent credentials in the electric generation industry.  The update included trade-off analyses based on the cost trends over the next ten years and how the mix of costs will impact the US energy picture.  Another topic was to identify supply chain issues in the wind energy sector so that the needed resources will be available to produce horizontal wind turbines.

The studies were all quite well done and very informative.  The top line result is this; in order to achieve 20% wind power by the year 2030 we will need to create 280,000 megawatts of new wind power.  If the projected average size machine is 2.6 megawatts, then that means 108,000 horizontal wind turbines will be needed over a ten year period.  That’s a lot of turbines.  And good news for GE and Siemens.  Maybe not so good for US workers if the majority of the turbine content is sourced outside the US.

At today’s cost for land based wind power, $2.47 million/megawatt, it will cost  $691.6 BILLION dollars over twenty years.  And if wind turbines actually produced their rated power, that might be OK.  But the fact is they do not.  The industry average is in the 22-28% range of productivity, with some hope of achieving 35% due to forecast improvements in the technology.  The utility factor takes into account the number of hours per year of wind blowing, down time losses, parasitic losses, etc.  The utility factor is also impacted by the poor system efficiency at anything other than the ideal wind speed.

If the electronics industry ran at this level they would have shut down years ago.   In response to this, the wind industry is moving offshore.  Why? Because offshore winds blow more frequently.  So the expectation is that by moving the wind turbines to the sea, the utilization factor will increase to 65%.  Great!

One small problem.  How do you make one of these things float?  Are they safe during a hurricane?   It’s like putting up a drilling platform, only harder because instead of drilling down, you have to go up with 20 tons of equipment suspended 250 feet above the platform.  The best estimates  for this feat of engineering put the cost at $5.94 million per megawatt.  Pushing the price tag for 20% wind power to $1.66 TRILLION.  And we haven’t begun to find out about the technical problems at sea compared to the problems we are discovering in land based wind power.

But here’s the scary part; you’ll never get to vote on it.  US energy policy is being implemented without Congress or legislative oversight.  The DOE has paid for a road map and by virtue of it’s $9 Billion per year budget, is spending that money and incentivizing businesses with matching grants.  R&D resources are being committed to achieve a goal that is, at best, very controversial.  We’ve had a lot of press, a lot of campaigning, but no real discussion and no real performance review.  But the DOE seems committed to the wind power roadmap, regardless of the cost or how long it takes.

Creating policy with your tax dollars.  With no input from you and me.  Now that’s scary!

Future Power 2010

Just visited the Renewable Energy World Conference.  Lots of really interesting topics to consider.  Lots of companies making progress in so many areas, it’s hard to sort everything out. Solar Power, tracking systems, Wind Power, engineering companies, geothermal systems.  All trying to define their role in the new energy economy.

Several companies, notably including Honeywell, are offering small wind packages.  These are systems with varying sizes of equipment that will generate electricity from wind.  Small wind can be defined as sized for operation of a single family home.  In the case of the Honeywell system it is designed to eliminate 18% of annual electrical power required.  With a target selling price around $4500 it will take a while for this unit to pay for itself.  And that is the difficult part about small wind.  There are many systems out there, from 2500 watts to 10,000 watts maximum output.  But the wind blows when it wants to, so it’s hard to tell what the payback will be for a small residential system.

A really “hot” topic is energy storage.  The demand for power does not follow the ability to produce power when wind or solar sources are being used.  So there has to be an intermediate storage capability to help the system manage the difference.  Anybody got a storage cell?

Well, as a matter of fact, A123 batteries who has been leading the lithium storage race, does have a storage cell.  They have a lot of them.  Some 2 Megawatt storage systems that fit inside a semi trailer.  Pretty impressive stuff.  They call it an APU, Auxilliary Power Unit.  Just like a diesel generator, only no smoke, no noise and no moving parts at all.  Pretty impressive.  And there’s more.

A123 has a technology migration path that they believe will create significant improvements over the currently available product.  The current wave of nanoscale processes applied to lithium battery chemistry will lead to decreasing costs and increasing energy density even further.  Both critical aspects of the business since competition for high energy density battery technology is coming from all over the world, including China and North Korea.  So if we want to hold onto this product technology and all the markets it applies to, we need to keep pushing the cost and performance envelope.

One interesting aspect of all this technology development is that we may see choices as consumers of power.  The current model of energy delivery being a highly centralized industry is going through a transformation of sorts.  This change may come quickly, and may have huge repercussions in our economy as the current group  of utility companies are forced to change their operating model.  What if everyone went off the grid?  What would the role of utility companies be? They might cease to exist, or become maintenance and operations experts keeping everything running.

What will the future of the power industry be?

More on Motor & Drive integration

The motor and drive combination is a basic building block of motion control.  Each component is useless without the other.  So it’s pretty important to come up with a good definition for what the “system” performance needs are to make sure that you end up where you need to be in your design.

Generally, we go to single source suppliers whose motors and controls are designed to operate together.  This approach guarantees minimum performance levels which the supplier can be held accountable for, and that’s a great thing.  But it usually comes with a huge list of features that will not be used in the application, but which you must pay for anyway.

In the industrial user community this is a great benefit because common environmental conditions such as dust, dirt, oil, or washdown conditions are more difficult to deal with and many products are built to these environmental standatds.  No special precautions are needed to make the equipment work.  And the suppliers warranty their equipment for continuous operation at stated speed, torque and environmental conditions.

But as we migrate into other businesses, machinery builders and equipment manufacturers have to apply the same motors and controls in larger numbers, and the overhead costs of motors and controls designed for industrial use become very expensive and make the cost of machinery less competitive.

In the machinery world, much of the electrical components will be built into control cabinets.  So if the packaging is going to be provided, is an enclosed controller somewhat redundant? Or can the motor speed control be integrated into the motor physically to reduce cost?  This is especially attractive if the cost of cabling a single motor can run 10-20% of the total cost of the motor and control.

There are performance elements that need to be considered in the applications as well.  How should the motor behave at zero speed?  What kind of torque regulation is needed?  What is the exact duty cycle of the application (on time versus off time)?   In these areas of motor and control operation the relationship of the motor to the drive is critical.  The drive electronics must have the ability to regulate current going through the motor without casusing overheating.  Some type of sensor on the motor must provide information about the position of the rotor at fairly high resolution, certainly more than every 120 degrees of rotation as with the common Hall effect sensors.

So as much as we may look at the purchase of a motor and controller as a system, there are a lot of nuances involved in the search process for the right hardware for any given application.  And that is probably one of the more difficult parts of doing motion control applications.

Motor and Drive Combinations

There is a subtle premise that often escapes us as we talk about motors and the controls that run them.  It is that the motor and controller operate as a package.  In most situations, a customer specification is for input voltage and output torque and speed.  That’s all that is important.  How you get there doesn’t matter a great deal.

But ironically, most motor manufacturers are predominately mechanical engineering centered.  And most drive electronics companies are electronics centered.  And they have very little in common with each other.  Except that their products must work together.  And oftentimes, that’s where the trouble starts.

The drive manufacturer warrants that his drive will produce current and voltage.  But the the motor can have very complex constraints to deal with in response to the excitation of the electronics.  How accurately a 6 step approximation of the sine wave performs, for example, can result in overheating in the motor depending on the loading of the system.  And as the motor winding heats up, the resistance in the motor can change dramatically, especially in the low inductance windings that are common in many specialty motors available today.

Then there are the cabling issues for connecting the motor and drive electronics.  The ac drive industry found out quickly that long wire runs can result in stored energy in the wires themselves.  Standing wave phenomena could cause higher voltages than expected and blow holes in the winding insulation in the motor.

Power semiconductor prices have fallen considerably in the last few years creating situations where it is sometimes cheaper and more reliable to put in parallel devices than to attached single power devices to large heat sinks.  This leads to some serious new options for packaging the electronics.  How about drive circuits in the end bell or junction box attached to the motor?  Actually, some models of the GE ECM motor (now owned by Beloit) are ac fan motors with variable frequency drives and intelligent controls built directly into the motor end bell.  You may have one in your main air handler in the air conditioning system of your home.  I was surprised to find out that I did.

I used to think that thermodynamics of these systems would be impossible to manage.  But the fact is that the drive efficiencies are getting really good.  One team I worked with was producing a 500 Watt brush drive that only shed about 20 Watts of loss at full load.  That’s some incredible efficiency.  So the notion of integrating motors and drive electronics is much more reasonable than it used to be.  And there are stepping motor packages that have been doing it for years.

So where is this all heading?

The fact is that the motor and drive electronics must work together as a package.  There is an increasing need, and an opportunity to create further performance enhancements, by the two technologies working more closely together.  More innovation will lead to better energy efficiency and new design opportunities and a chance to recharge (pun intended) an industry that has been losing share to offshore competition in the last few years.

Wind Energy Equipment Testing

Some friends were discussing the recent visit of Department of Energy Secretary Steven Chiu to Clemson University to deliver a check for $45 million to start a test facility for horizontal wind turbine gearboxes.  It seems that there have been a number of gearbox failures in recent years that suggest a wider problem that will need to be solved in order for wind energy to become as reliable and cost effective as promised.  Gear boxes are failing in the range of 10 years of operation, and occasionally, sooner than that.

This is not difficult to understand.  The forces acting on the gearbox are huge.  On the input side you have 3 blade propeller with blades that are now approaching 200 feet in length.  I don’t care how light weight they are, carbon fiber epoxy or Kevlar or whatever, the forces are tremendous.  In addition the blades have to rotate to take them out of the wind when the wind is too fast for the system to operate.  So there are actuators at the base of the blades adding to the weight and mechanical complexity.

Then there is the intermittency of the wind itself.  This can manifest itself as bursts of wind or winds of different speeds hitting the same rotor.  Which can lead to all kinds of impulse loads on the gearbox.  Gear teeth becoming momentarily unloaded and loaded in response to the wind.  This is actually one of my favorite “Stump the Band” questions for mechanical engineers; what is the formula for the shock load of gear tooth reversal?  It’s big, whatever it is.  And the shock load of the propellers is driving the gearbox against a high inertia load, the generator.  So there is a lot of resistance to overcome.

But the really scary part is that the gear systems are often in the range of 30,000 pounds in weight.  And they are mounted on metal masts at heights of 1.5 times the blade length.  So that would be 300 feet up in the air in the case of a system with a 200 foot blade.  Making the replacement of a failed gearbox a bit more complex than dropping the transmission out of a car, for example.  Especially since most wind farms are in very remote locations where the land is cheap and the wind blows some of the time.

This lead the Department of Energy to put out requests for proposals to address the technical question of providing the industry with a resource to help in the design of gearbox systems with much higher reliability than the current designs.  Total cost of this effort, approximately $100 million dollars.  The proposed test facility is targeting 20 megawatt power handling capability, or approximatley 27,000 horsepower depending on the exact rpm of the system.  This is an incredibly big piece of machinery.

Clearly, gearbox technology has to get better for the wind industry to continue to prosper.   I wonder if we are putting a band aid on a technology that is fundamentally flawed.  Maybe we need to be concentrating on the next generation of the technology and improving the cost performance by an order of magnitude.  Surely we can do better.

New Motion Feedback

The field of motion control is heavily dependent on the feedback device.  There are a world of issues encompassed in this statement.  But I will skip the majority of them and jump to the conclusion that magnetic encoder technology has long been a favorite of mine as a potentially ideal solution for a number of reasons.  Magnetic position technology tends to be more resistant to environmental problems such as temperature, dust, dirt and humidity.  Since all motors are heat producing systems, temperature restrictions for feedback technology can be a problem.  And since electric motors are frequently found in environments where there is dust, dirt and humidity, magnetic feedback would be ideal.

On the other hand, magnetic feedback has been complex and expensive in the past.  Resolvers require high precision windings in the sensor and precision power supplies to excite them.  Can you spell “expensive”?

Enter the Hall sensor.  In spite of the fact that the Hall Effect has been understood since 1879, the use of the technology has only recently become widespread with the fabrication of semiconductor level Hall devices.  The Hall sensor as a transistor found very popular application in sensing the three phases of brushless dc motors’ permanent magnet rotor.  The bldc motor technology was essentially impossible without this crucial piece of technology because in the early versions of the control, it was impossible to start the motor without knowing which phase to energize.  This has been less of a problem with the advent of low cost, high performance microprocessor controls that are able to run brushless dc motors with or without Hall sensors.

New arrangements of the Hall devices into arrays with greater capabilities is where the Hall effect technology intersects the position feedback technology.  The Hall arrays are capable of sensing small permanent magnet domains on rings that permit rotorary position to be sensed in either analog or digital form.

While there are a number of suppliers of Hall sensing arrays for motion control, a couple of new twists have been added.  The Timken company has added some new features to the Hall array that have additional benefits.  Among them are the ability to program the numerical value of the digital output, which can be a very helpful feature that eliminates fractional remainders and rollover error in control systems.

In addition, Timken is introducing a new linear version of the technology which is a real first for the motion community.  Most linear motion is the result of converting the rotary motion through a linear mechanical device, either a belt or leadscrew.  But the control system is measuring the position from the feedback  on the motor that’s driving the system and not on the load.  So the mechanical error of the leadscrew or belt is the limiting factor in high performance linear.

And to make the new magnetic feedback really interesting, it is comparable in cost to conventional optical encoders.  Which is really going to create some new opportunities for everyone in the motion control field.

Innovation and Growth in Robotics

The robot industry has gone through some interesting changes over the years.  Most of the companies that were involved in the start of the real robot revolution are gone, unable to meet the extraordinary cost reductions that were sure tocome in order to make robots cost effective in most industries.  The biggest lesson, in my opinion, was the idea that robots had to be narrowly defined in terms of their application.   There was a time where there were only a few companies with the control technology to be able to make the multi-axis coordination work correctly.  So every application had to be programmed from scratch and the learning curve was huge.

The fact is that a welding robot is nothing like a Cartesian robot for electronic assembly.  And part of the learning curve of the industry was understanding what applications to focus on.  This first big reality set in when many companies began to compete for welding applications because the automotive market  opportunity was huge.  And just figuring out one application was a big enough task that it consumed most of the development resources available in  companies like GE and ABB robotics.

Consider the huge learning curve that has taken place in 35 years.  Medical robots have matured to the point where orthopedic surgery by a robot is faster and more precise than the best surgeons.  Researching the human genome would have been impossible without the high speed sample management systems of bio-assay robots.  And autonomous robots have searched the inside of volcanoes, taken samples on the moon and roamed and photographed Mars.  Pretty impressive.

Consider the forecast for the future of robotics. Motors and controls have become incredibly sophisticated and costs have dropped dramatically.   Computing power has increased to the point where memory and processing costs are almost trivial.  The First Robotics Competition is bringing 150,000 school children into the field of robotics through its programs with schools all over the US.  And the knowledge base and experience is so pervasive that we have Lego making teaching systems for grade school children to begin to get exposure to robotics.

Among the amazing developments, Barrett Technology has an anthropomorphic arm and “hand” gripper that is designed to low force, low power consumption and safe enough to be in proximity to humans.  The Robots and Mechanisms Lab at Virginia Polytechnic has demonstrated many new solutions to common problems of robot locomotion culminating in the Darwin soccer playing robot that operates autonomously.  Their goal?  Team Darwin wants to be able to compete with human soccer players by the year 2050.

With this kind of innovation, the future of robotics is going to be great.

Magnetics 2010 and Motion, Drive & Automation

There is a small industry conference that takes place every year with a lineup of industry experts that is top notch by any standard.  It’s called the Motion, Drive and Automation Conference put on by E-Drive magazine.  This year it is located at the Disney Hilton Resort in Orlando and is taking place on January 28 & 29.   The conference includes a wide range of industry experts from many fields of advanced electric motor design, advanced motor control concepts, power semiconductors and state of the art motor testing system.  There will be a lot of technical and product presentations that showcase leading edge technology in electric motors, precision gear reducers, new technology for motion sensing, and a number of improved power semiconductor devices for the motor control industry.  This is a great place to get up to date on the latest technology that will impact of motor and control technology across many industries over the next few years.

In addition, the Magnetics 2010 Conference will be running concurrently at the same venue.  Magnets are a strategic material without which many motors would simply not operate.  In the ever-changing motor industry, there is always a new design that seeks to make an enhancement over previous solutions, or introduce a new solution to old problems.  Declining prices for Neodymium Iron Boron magnets over the last few years have created a number of novel design shifts which have been instrumental in bringing more varieties of permanent magnet machines into the forefront of motion control and mechatronic technology.  To the point where over the last two years a resurgance of permanent magnet rotor designs have been created to improve the energy denisty and lower the cost of specialty motors in washing machines and air conditioning compressors.

This last development, combined with the forecast increase of hybrid electric car sales coming this year, are expected to increase the sale of permanent magnets by 10-15 percent by 2011.  That’s a staggering jump in a market that is almost exclusively supplied by China.  And there is no assurance that China can meet the forecast production.

The US Department of Commerce usually has a say in the sale of products or businesses to foreign countries that are deemed to be strategic or sensitive technology.  In fact, I got stuck in a situation where my employer was told specifically that we could not sell a CNC controller to a Korean customer.  That’s pretty small potatoes compared to controlling the supply of permanent magnets which influences billions of dollars worth of electric motors manufactured and sold all over the world.  So it strikes me as a little odd that the sale of Magnequench to its current owners, Neo Materials, was completed without a much discussion. leaving the US without a domestic magnet supplier.

There will surely be a lot of discussion about this situation at the conference, and I will be in attendance to get the latest information on the subject.  So look forward to a review of the conference in an upcoming post.

Inventing Industry in the (near) Future

The future of the US economy, and our future as an industrial power will be the result of our cumulative creativity.  New industries will be the result of new ideas, new technologies, new thinking.  It’s gratifying to see programs like the First Robotic Competition getting 215,000 junior high school and high school students exposed to and involved in robotics.  Problem solving, finding solutions, getting their creativity flowing to make a box of parts into a working machine with real world performance.  It will be even more interesting to see what those same kids will be into 5 to 10 years from now as they begin their careers in the many technology pursuits they are likely to follow.

Technology is a major driving force in the economy.  The ability to create whole new industries that have never existed before.

And there is a second driving force, sometimes made less obvious by the flash of the latest technical breakthrough.  Cost.  What is the relationship of cost to the development of industry?  As costs decline volume goes up.  Steel manufacturing per man year of labor increased 500% during a period of intense competition between the US and Japan.  And interestingly, one of the breakthroughs was the creation of the “mini-mill” which could produce specialty steels more cost effectively by making them in smaller batches.  Sometimes the solution is counter intuitive.  The steel industry was all about increasing batch size.  But serving the market with more complex products turned out to be easier with smaller batches, ultimately increasing overall sales and defending the US market to some extent from foreign competition.

Are there other cases where innovation was economically driven?  In the machine tool world the majority of manufacturers develop bigger and more complex machines so that a single machine can handle any operation.  This complexity tends to drive costs up quickly.  So the tendency is to find high performance machine tools costing hundreds of thousands of dollars.  In contrast, the HAAS company re-invented the machine tool business by focusing on making a low cost, high quality machine tool that many shops could afford to buy.  They were one of the first companies to have several models of machine tool in the $50K range.

They did it by concentrating on the economics of a machine tool that was profitable in operation.  That means a machine with a low cost to purchase, low operating and maintenance costs, and sufficient precision to meet the requirements of most operations.   In order to reduce their machine cost they had to develop their own controls platform.  They restructured everything in the design and manufacture of the CNC system to meet the cost objective.

In act, they are so successful, that HAAS is the largest CNC company in the western world.

Many similar situations exist in other industries.  In small plastic parts manufacturing there are a number of breakthroughs that have created lower cost parts in smaller batches based on innovative new tooling systems.  In metal fabrication there are new process like thixotropic molding and metal injection molding that have been developed to lower the cost of metal goods by making parts at lower costs.  These solutions are focused on reducing costs and other barriers to the entre of new products like tooling costs and minimum batch sizes.  And they represent major new markets that were not possible in the past, because they are focused on the economics of the industry they serve.  Decreasing the cost of entry and the cost of part manufacturing opens up new markets

So inventing the future can be technology.  Or as it can be economics.   It’s all innovation.  And it’s all about delivering value.

Next Page »