We are saddened to report that Professor Dornfeld passed away in March, 2016. If you enjoyed his blog, please consider making a contribution to either of two funds at UC-Berkeley that have been established in his memory.

David A. Dornfeld Graduate Fellowship
David A. Dornfeld Scholarship

Saturday, August 27, 2011

Less is more, part 5


Designing for small

In the last posting we stated that one of the challenges is linking product performance to material shape and properties. And then making the next link to environmental impacts/resource requirements.
An example of some helpful software that connects material properties to potential environmental impacts/burdens was given. By linking the potential burdens to material properties, and then to the design or production requirements, we can try to choose the least impactful material.

So, with respect to either a process/machine for manufacturing (manufacturing phase) or product (use phase), the challenge is to find the design/material/structure combination that:
i. gives the desired performance/meets specifications
ii. can be economically manufactured/operated at sufficient scale with required production rate, quality, and cost,
iii. while minimizing the environmental impact or, better, reducing the impact enough compared to the present performance to offer a "return on investment" that moves the operation of the process or product towards a more sustainable situation.

One of Ashby's techniques to start his analysis (see last posting for more info on Ashby!) is the "use matrix." This matrix arrays, vertically, energy intensive to material intensive products and, vertically, different product "load factors" from high impact to low impact. For example, the categories of energy to material intensity are from primary power consuming to non-power consuming. The primary power consuming products are energy intensive in their use phase and the non-power consuming are, then, material intensive. The figure below is reproduced from Ashby's CES Eco-selector white paper from February 2005.


In this matrix you can see examples of entries in the various categories with an automobile being a primary power consuming high load factor product (meaning the use phase impact is power related) while on the other end of the scale a tent or canoe is low load factor and material intensive since the tent requires no energy to operate so the material consumption in the manufacturing phase is the most significant.

Now, if we looked at manufacturing in the same manner, what could be a “use matrix of manufacturing classes”? Here is my attempt to fill in such a matrix for manufacturing.


You can follow the logic I think. An example of a high load factor energy intensive manufacturing process is something like a furnace for heat treating or a semiconductor manufacturing etch tool. A low load factor manufacturing process or element could be a warehouse or office for a factory which is midway between energy and material intensive depending on the exact activity in that warehouse or office.

Ashby's process uses such a matrix to help determine which phase of the "product" (here a consumer product but in our discussion a piece of manufacturing hardware for a process or factory component),  that is material production, product manufacture, product use or product disposal/end of lied, should be focussed on for the largest improvement.

If one is designing or producing high or modest load factor primary power consuming machines for production, such as rolling mills, forming presses or machine tools, etc. as in the manufacturing matrix above, then we would want to consider these four phases relative to those machines.

Let's consider the example of the design of a deep draw press. We'd like to come up with a press that meets the constraints posed at the beginning of this posting - gives the desired performance/meets specifications, can be economically manufactured/operated at sufficient scale with required production rate, quality, and cost, and minimizes the environmental impact.

If you are not sure what these are there is an excellent on-line video on the operation of one made in Taiwan and its manufacture. Note: this is a sales video but informative! The process performed on such a press is more objectively detailed on Wikipedia under deep drawing.

The elements to be considered in the design of a deep drawing press would include:

- Material production: steel mostly (several tons)
- Manufacture: welding (mostly), machining (some), electronics (not many)
- Use: electricity, hydraulic fluid, compressed air and other consumables
- Disposal: scrap (likely sold for re-use)

The design criteria would include:
- tonnage (pressure/power) which determines the size of the part to be made or thickness of the metal formed
- stiffness
- speed/strokes per minute
- ease of load/unload
- die changing/handling/setup

The press capacity is determined by the tonnage it provides for deep drawing while maintaining the necessary stiffness for the accuracy of the forming process. The speed is dependent on the efficiency of the energy to move the press given the weight of its components. A press that move rapidly (up/down strokes) either must be light (and hence low tonnage) or require a lot of energy to move.

Ashby data provides a measure of the relative "cost" in embedded energy of different materials per unit bending stiffness (affecting precision) and mass per unit of bending stiffness (for the speed vs precision tradeoff).

The curve below, from Ashby's software, shows the "trade-off surface" for this energy-mass for a stiffness limited design. The curve shows the range of reasonable candidate materials for achieving the required mass (for speed) and stiffness (for accuracy) normalized by embedded energy. Ideally, following along this curve gives the designer a set of material that will meet these constraints.


We see that one of the materials lying near the curve is cast iron and another is mild steel in the lower right part of the curve- both reasonable cost alternatives. Others on the curve, but with higher cost, are beryllium alloys in the upper left part of the curve- not likely to be used. Also not likely to be used is chipboard which is a bit below the curve. Another material not traditional used but worth considering is carbon fiber reinforced plastic - one the curve near the bend. These fiber-based materials offer very high strength/stiffness and very low mass so could be a new design for presses for high speed but high stiffness with similar embedded energy, for the amount needed, as steel.

These materials (the steel and cast iron at least) are also easily recovered at the end of life and, in fact, lend themselves to re-manufacturing (another good topic we'll delve into sometime) as well.

Next time we'll apply this to the manufacture of a precision machine tool.

Monday, August 8, 2011

Less is more, part 4


How much less is less?

The last several postings have been discussing the elements of reducing consumption in manufacturing. Not just cutting but making better use of the resources available. This stretched from reviewing the "buy-to-fly" ratio concept to yield issues in metal production and use. We discovered that there is a lot more potential in the material that is left on the floor in production than we might think. In fact, improving material processing yield may actually offer more potential for impact reduction than many other strategies.

But these are technically complex issues. Manufacturers don't waste material on purpose. The swarf from machining is due to material removed to achieve the desired shape. The farther the input workpiece is from the final shape the more material must be removed and shows up as chips on the floor. These chips are routinely recycled of course. But that is a far cry from not using it in the first place.

The term "net-shape processing" (defined as making things to a final or near final shape without removing material - such as forging) is one approach to reducing the amount of material that needs to be removed to achieve the final shape. This cannot address the requirements of surface conditions (like very low roughness) or some form requirements but it goes a long way. This does not work for all materials. But, for example, plastic injection molding is a classic example of net shape forming (except for the runners, sprue, etc. unless done with hot runner systems as in high production.)

The challenge is linking performance to shape and properties. And then making the next link to environmental impacts/resource requirements.

Engineers like to use "tools" for assisting in making these links. By tools we mean software or other analysis methodologies that assist in presenting data or alternatives to the designer, or manufacturing engineer, to be used in decision making. These tools often help the engineer answer questions like

- what is the function of the device or piece of hardware or component that is being designed?
- what are the objectives that need to be optimized?
- what constraints must be satisfied?

These questions are common to all engineering design problems but are part of the concept behind a wonderful tool from Granta Design called the CES (for Cambridge Engineering Selector) methodology developed by Professor Mike Ashby and his colleagues at Cambridge University in England. This was first developed to help engineers and designers to select materials for use in products and components.

They give the example of a design relative to these questions as "For instance, a car body panel (function) needs to be as light as possible (objective) for a specified stiffness and cost (constraint). Other constraints on the design might be acceptable resistance to mechanical impact and to contact with various environments." This is described in great detail on the Granta website.

I need to mention here that I am in no way associated with Granta Design or Mike Ashby and am not being paid to pitch his product or company! We are, in my lab, using this software (and we paid for a license) and I am only a big supporter because it is one of those products that is very useful and enables us to do things we otherwise would not be able to do. I also use Granta software in my class on sustainable manufacturing.

Ashby developed the concept of selection charts that show one type of material property as a function of another - for example elasticity as a function of thermal distortion. So if you are designing a component, say of a machine tool, and you need a material that has a certain stiffness but is less sensitive to temperature variations (and the accompanying distortion, growth with increase in temperature and shrinkage with reduced temperature) then you could see where different material groups fall and choose a material that is in the range desired.

The data is the same data you'd get from a handbook, or tests, or another expert but the method of presenting it offers additional insight to the designer.

For example, the figure below, from the CES EduPack Manual from Granta (this is available on the Granta's website for teaching tools) shows a typical "Ashby chart" plotting Young's modulus (this is the "elastic limit" for materials in load per unit area which serves as a measure of the stiffness of an elastic material) as a function of density (mass/volume).


This would be the kind of information that an auto designer would use to pick a material that has the required stiffness but with the least possible weight (since lighter vehicles require less fuel to move and lighter frames to support them).

Ashby fits the use of this data in assessing material selection impacts over the life cycle of a product. The figure below, also from Ashby and available at the link above, shows the stages of material usage in a product lifecycle. Ashby data makes it easier for an engineer to see



the magnitude of these impacts. And, of course, it gets back to our "make vs use" impact discussion some time ago (for example as part of the material diet discussion). The take away from this figure is that there are materials issues at all stages of product life.

The one limiting element of Ashby is that he looks at product design through the lens of materials and there are, of course, other concerns. But, this is a small issue compared to the benefit of his approach.

So, where does the "green-stuff" come in?

One of the axes of information that Ashby provides is embodied energy (and its equivalent in CO2 emissions). The figure below, also from Ashby, shows embodied energy (GJ/volume) for a wide array of materials. This data, specially when plotted as a function of material parameters, opens up



the possibility of connecting design parameters to environmental impact. The embodied energy data Ashby relies on is generally reliable. In a number of cases the data may reflect a specific region of the world or particular means for processing the material but it is an excellent base to work from. The red dotted line is only for comparison of materials and embodied energy.

As you can see in the figure above, the potential for looking at energy (and thus CO2) impact for materials is wide open. Within a materials group, like polymers, there is a factor of 10 range of embodied energy. Within the ceramics group this range grows to a factor of 1000 (three orders of magnitude.)

Clearly, all these materials in each group do not have the same properties are, hence, are not interchangeable. But, linking them to material properties, then to the design or production requirements, lets chose the best, least impactful, material.

That's where we start next time and we'll include an example.