--------------------Comment by Dr. W. Craig Carter----------------

Dear Prof. Kim,

Thank you once again for organizaing such a delightful and informative workshop.  I hope the following comments do not arrive too late.

Random Comments

1.  The idea of the workshop was very interesting. The realization that there is a large intellectual disparity between fields mechanics-mathematics-materials science on which topics are pertinent for nanoscale science made a large impact on me.  Not only was there confusion on the pertinent scales which must be considered, there is little agreement on appropriate techniques and levels of rigor which must be applied to nanomechanical behavior.

2. I think it is important to understand how techniques are implemented at various scales.  It would be useful if the challenges which come from crossing between scales could be alleviated by agreed formats for data exchange, an attempt to collect results and convey them to related fields, and whenever possible produce example software which allows others to adopt techniques across scales.

3. Candor must always be applied on what a particular technique can and cannot do considering reasonable computational resources.  This should be considered as part of the scientific ettiquete and necessary for useful communication between scales. When new techniques are developed, comparisons to existing techniques (if they exist) to do the same types of problems should always be presented.

4. The polymer and biomaterials communities present us with interesting challenges which focus on complex molecules and their interactions.  We should not attempt to oversimplify these problems when we work on them if we intend to make serious progress.

5. When we begin investigations in a new field, it is appropriate to carefully consider what progress has been made, what is accepted as the standard models and credit those who have made progress.

6. We need to distinguish between those phenomena which truly do not scale with size in an expected way from those for which boundary conditions begin to have a large influence on the solution. Scale-dependent constitutive properties should not be confused with boundary condition dependent properties.

7. The theory and modelling researchers need to make an effort to understand the meaning and interpretation of new materials characterization.

8. The mathematicians have a lot to offer us, we ought to pay more attention to what they are saying.
--
W. Craig Carter
Lord Foundation Associate Professor of Materials Science and Engineering
MIT, Dept. of Materials Science and Engineering
13-5034  77 Massachusetts Ave, Cambridge, MA 02139-4307 USA
617-253-6048  FAX:617-258-7874 [email protected]
http://pruffle.mit.edu/~ccarter
schedule--------> http://pruffle.mit.edu/~ccarter/schedule.html
--------------------------------------------------------------------------

-------------------Comment by Dr. R.W. Carpick-----------------------
Dear Kyung-Suk,

Thank you again for the opportunity to attend and present at the NSF workshop. I learned quite a great deal, and it was especially useful for me to make contact with people in the mechanics community. I am writing you with some feedback and ideas regarding the topics that were discussed.

I was struck by the fact that several studies, both theory and experiment, indicate that continuum mechanics still 'works' at the nanometer scale. My own friction results are consistent with this, but also Rob Philips' work, David Srolovitz, Pharr's results, and others, indicate that in a variety of situations, continuum mechanics can be consistently applied. Still, we cannot use continuum mechanics to extract materials parameters - only experiment or atomistic theories tell us this. The relevance is that continuum mechanics can be applied to experimental results to extract materials parameters, such as the interfacial shear strength in an AFM friction measurement, for example. Coming from outside the mechanics community, I can tell you that many other researchers do not appreciate this. Typical physicicists or chemists doing AFM measurements, for example, rarely know how to apply continuum mechanics models to even begin to describe their results. Either they do not believe it will work, or they are not familiar with it to begin. The mechanics community has a lot to contribute here; interdisciplinary research involving mechanics people to help solve research problems in other domains would be valuable.

Furthermore, there were also examples where deviation from continuum behavior was evident; Huajian Gao had an example of this. Likewise, the atomic stick-slip behavior I observe is, by its very nature, a non-continuum phenomenon. The question that emerges for me is: when can I safely apply continuum mechanics to model my experiment? Put another way: how far can we take continuum mechanics without serious modifications? It would seem that we need further direct comparisons between atomistic theories and continuum mechanics, with experiments to back up these comparisons. I'm speaking very generally; this includes contact mechanics, fracture mechanics, elasticity, thin film properties mechanical properties etc.  Studies that test the validitiy of continuum approaches at this new, small scale of interest should be supported. This would include both experiment and theoretical comparisons.

Pushing theoretical and experimental mechanics to smaller scales is absolutely necessary when studying novel materials problems, namely problems involving nano-scale materials (nanotubes, nanowhiskers etc.) and biological moleucular materials. For example, the incredibly high strength of nanotubes and nanowhiskers could presumably be used to great advanatage for improving fiber-reinforced composites. Since these materials are only a few nanometers in width, we cannot use conventional tools for studying their inherent mechanical properties. This is one of many examples that indicate to me that it is important to support instrumentaldevelpment that enables mechanical measurements at the nano-scale. The atomic force microscope is a powerful tool, but there is much more work to be done to improve it and develop new instruments for these purposes.

I hope these thoughts are useful, and I would be happy to discuss them with you in further detail. When do you expect to generate a draft report? I am very interested in reading it and providing further feedback.

best regards,
Rob

P.S. I have one more comment to add regarding the NSF workshop, if it is not too late. As we shrink the scale of the materials and devices we are investigating, the larger surface-to-volume ratio will increasingly force us to deal with surface/interface issues. I would suggest that the dreams of "nanotechnology" will not be realized if a better understanding of interfacial adhesion and friction cannot be developed. The control of interfacial forces will be necessary to allow the manipulation and assembly of nano-componenents, for example. This is a wide-open area, and clearly mechanics can and should play a large role.

----------------------------------------------------
Dr. Robert W. Carpick, Sandia National Laboratories

http://www.sandia.gov/surface_science/nsom/nsom.htm
phone: 505-844-9194 fax: 505-844-5470
email: [email protected]

US Mail: Sandia National Laboratories, MS 1413, Albuquerque, NM, 87185-1413

FedEx, UPS etc:  Bldg. 897, Rm. 2012, 1515 Eubank St. SE, Albuquerque, NM,
87123-1413
-----------------------------------------------------------------------------------------

---------------------Comment by Dr. Yonggang Huang---------------------------------

Even though no one has explicitly defined what nano-mechanics is, the title is self-explained -- mechanics at the nano-scale.  At this scale, the boundary between mechanics and physics is not so clear, but nano-mechanics is still unique and is complementary to physics.

Molecular dynamics (MD) and other atomistic and quantum simulations in physics certainly play an important role in understanding the material behavior at the nano-scale.  However, these physics-based simulations have limitations.  For example, the time scale in molecular dynamics is on the order of pico-second, several orders of magnitude smaller than any time scale in real life.  Moreover, MD simulations tend to generate tons of data that may overwhelm some key physical parameters.  It is therefore important to have a nano-scale continuum mechanics theory to characterize the collective behavior of materials; to extract important information from physics-based simulations; to identify the key physical parameters that govern the material behavior at the nano-scale; and to guide further extensive MD simulations.  One example has appeared in the study of carbon nano-tubes.  The MD simulation can indeed simulate the deformation pattern of single-wall or multiple-wall carbon nano-tubes.  However, in order to extract key parameters governing the deformation of nano-tubes (such as the critical strain triggering different deformation modes), an elastic continuum theory must be used in conjunction with MD simulations (e.g., Yakobson et al., 1996; Lu, 1997; Wong et al., 1997; Yakobson, 1998).  This clearly shows that the continuum analysis can indeed capture the deformation modes of nano-tubes observed in MD simulations, though the former is much simpler and more efficient.  Therefore, a continuum mechanics at the nano-scale, in conjunction with the MD simulation, is important to nano-scale technology and science.

There are many important mechanics issues at the nano-scale.  Several are discussed in the following.

(1) Development of physics-based nano-scale continuum mechanics theories
Quite a few speakers in this workshop have compared the MD and other atomistic simulations with the theory of elasticity, and have concluded that the elasticity theory can characterize the deformation rather accurately down to a few atomic spacings.  This conclusion, however, is primarily based on the comparisons of MD simulations with the theory of linear elasticity since the materials and structures examined in these studies undergo very small deformations.  In the theory of linear elasticity, the only parts that interact with the material behavior at the nano-scale are the elastic modulus, which can be readily obtained from the MD calculations.  Therefore, under small deformations, the MD simulations and continuum mechanics theories are bridged at the level of linear elasticity constants, which is relatively simple and straightforward.
The experiments and MD simulations at the nano-scale, however, have clearly indicated that some nano-scale materials and structures undergo very large deformations.  For example, the strain levels observed in compression, bending and twisting of nano-tubes, nano-rods and nano-ropes are typically on the order of 6-10% (Yakobson et al., 1996; Lu, 1997; Wong et al., 1997; Yakobson, 1998).  In fact, that nano-tubes being capable of sustaining large deformation without failure are the basis for many proposed applications of nano-tubes ranging from nano-composites to microscopy, and are important to the nano-scale technology.  The linear elasticity theory, however, becomes invalid at large deformations.  Instabilities in the form of bifurcation, buckling, and necking, which have been repeatedly observed in the nano-structures (Yakobson, et al., 1996, 1997; Lu, 1997; Wong et al., 1997; Lourie and Wagner, 1998; Yakobson, 1998), also result from the nonlinear deformation at the nano-scale and cannot be characterized by a linear elasticity theory.  Therefore, a nano-scale nonlinear continuum theory that can represent the effect of finite deformation needs to be developed.
It should be pointed out that it is much more difficult to develop a nonlinear continuum theory at the nano-scale than the linear elasticity theory because the construction of a continuum theory from nano-scale physics is certainly much more challenging than the calculation of linear elasticity constants from MD simulations.  In fact, even though there exist many phenomenological nonlinear elasticity theories at the macro-scale, they are not applicable at the nano-scale since they are not physics-based theories.  One challenge in nano-mechanics is to develop a nonlinear continuum mechanics theory based on nano-scale physics, such as the inter-atomic potentials or other cohesive laws of materials, or MD and other atomistic simulations.  The establishment of such physics-based continuum mechanics theory exemplifies the nature of nano-mechanics – bridging continuum mechanics with physics at the nano-scale.

(2) Development of failure mechanics theories at the nano-scale
The proposed applications of nano-tubes range from nano-composites to microscopy since they are believed to be the strongest filaments (Wong et al., 1997; Yakobson, 1998).  Even though the nano-tubes have relatively large strength, they eventually fail when the deformation reaches some critical state.  But the failure modes of nano-structures are very different from those at the macro-scale, and cannot be characterized by the macro-scale failure theories.  For example, recent simulations and transmission electron microscope (TEM) observations have indicated the carbon nano-tubes elastically buckle (versus fracture and plastically deform) when bent to large angles (Ebbesen and Ajayan, 1992; Despres et al., 1995; Iijima et al., 1996; Yakonson et al., 1996; Cornwell and Wille, 1997).  On the other hand, the carbon nano-tubes under tension have a very large breaking strain and, at the later stages of fracture, the nano-tubes fragments are connected by a set of unraveling monoatomic chains.  These carbon nano-tubes under tension, however, can be relaxed by creating a dislocation dipole, which leads to a nano-tube of smaller diameter and therefore significantly changes the electrical properties.
The application of nano-tubes requires an accurate assessment of the potential failure modes.  Unfortunately, classical failure mechanics theories, such as fracture mechanics, are not applicable at the nano-scale.  The basic concepts established in fracture mechanics such as the stress intensity factor K become invalid for nano-structures, which are too small to have a K-dominance zone.  Even though the concept of “toughness” is still being discussed for nano-tubes (e.g., Wong et al., 1997), there are no established failure criteria in nano-mechanics.  In order to accurately assess the reliability of nano-structures, it is important to develop new failure theories based on physics at the nano-scale, such as the inter-atomic potentials or other cohesive laws of materials, or MD and other atomistic simulations.  It should be emphasized that, in the atomistic simulations, no separate failure criteria are imposed such that fracture and failure of materials become a coherent part of deformation analysis.  In nano-mechanics, it will be desirable that the failure theories also become a coherent part of the constitutive theory such that no separate, adhoc failure analysis is needed.

(3) Multi-scale analysis and scaling laws
Several speakers at the workshop have pointed out that the interactions among multiple scales are important to the success of nano-scale science and technology.  In particular, the scaling laws can play a critical role in nano-mechanics.
It is clear that different research communities (e.g., physics, mechanics, engineering) may have quite different interests in nano-mechanics.  For example, physicists are interested in the atomistic and quantum simulations where the relevant length scale and time scale are typically on the order of nano-meters and pico-seconds, respectively.  The mechanics people, on the other hand, are interested in the continuum analysis where the corresponding length and time scales are generally above ten nano-meters and nano-seconds.  It is quite a challenge to link these different theories over such a large span of scales.
The development of scaling laws represents an important step in linking different scales.  If a scaling law can be established over certain scales (e.g., > 100 nano-meters) with one or several scaling coefficients to be determined, researchers interested in the next scale (e.g., < 10 nano-meters) can then focus on the determination of these scaling coefficients by the corresponding physics laws at the next scale.  The researchers at the small scale do not need to perform the nearly impossible (if not impossible) full-scale simulation over the large scale, while the difficulty of determining some scaling coefficients the researchers face at the large scale can be alleviated by the small scale computations or experiments.  This may provide an effective way to connect different scales as well as different research communities together in the development of nano-mechanics.

The development of nano-mechanics represents efforts that must involve multi-scale and multi-disciplinary approaches.  As pointed out by several speakers at the workshop, linking length scales is critical to the success of nano-mechanics because each disciplinary has its limitations.  The interactions among different disciplines (physics, mechanics, engineering) can help us to gain insights of the nano-scale science and technology; to address cross-disciplinary problems; to identify critical issues at each scale; to determine the key parameters that cannot be obtained at the samescale; and even to develop new theories based on the physical laws at the next scale.

Reference
Cornwell CF and Wille LT, Solid State Commun., v 101, p 555, 1997.
Depres JF, Daguerre E, and Lafdi K, Cardon, v 33, p 87, 1995.
Ebbesen TW and Ajayan PM, Nature, v 358, p 220, 1992.
Iijima S, Brabec CJ, Maiti A, and Bernholc J, J. Chem. Phys., v 104, p 2089, 1996.
Lourie O and Wagner HD, “Transmission electron microscopy observations of fracture of single-wall carbon nanotubes under axial tension,” Applied Physics Letters, v 73, pp 3527-3529, 1998.
Lu JP, “Elastic properties of carbon nanotubes and nanoropes,” Physical Review Letters, v 79, pp1297-1300, 1997.
Wong EW, Sheehan PE, and Lieber CM, “Nanobeam mechanics: Elasticity, strength, and toughness of nanorods and nanotubes,” Science, v 277, pp 1971-1975, 1997.
Yakobson BI, “Mechanical relaxation and intramolecular plasticity in carbon nanotubes,” Applied Physics Letters, v 72, pp 918-920, 1998.
Yakobson BI, Brabec CJ, and Bernholc J, “Nanomechanics of carbon tubes: instabilities beyond linear response,” Physical Review Letters, v 76, pp 2511-2514, 1996.
Yakobson BI, Campbell MP, Brabec CJ, and Bernholc J, “High strain rate fracture and C-chain unraveling in carbon nanotubes,” Computational Materials Science, v 8, pp 341-348, 1997.

-----
Professor Yonggang Huang
Department of Mechanical and Industrial Engineering
University of Illinois at Urbana-Champaign
140 Mechanical Engineering Building, MC-244
1206 West Green Street, Urbana, IL 16801

Telephone: 217-265-5072
Fax: 217-244-6534
Email: [email protected]
---------------------------------------------------------------------------------------

-------------------Comment by Dr. Shefford P. Baker--------------------------------

Dear Kyung-suk,

Sorry for the delay in getting my comments to you regarding the workshop in Palo Alto. I found it to be a very interesting workshop and a very good group of people. Thank you for organizing it. I think that periodic meetings like this to try to identify where we should (and should not be) going in this field are very worthwhile. I have only a few remarks to add:

During the summary discussion, most folks seemed to be saying that goals in this field are much more complicated than in, say, Biotechnology. I don't think so. I think we have a simple premise -- that things in small dimensions don't behave like things in large dimensions. Therefore, they are interesting to study. I think that the projected outcomes are also fairly simple: Scientifically, by understanding behavior in constrained systems, we learn not only about those constrained systems, but also a lot about larger systems. Technologically, since capability/unit volume represents "power" density, the ability to do more in a smaller volume almost certainly will lead to new wonders. Finally, I think that we pretty much agree on what we need to focus on. We need to integrate our many disparate experimental and model approaches. Learning how to bridge the gaps between them, or at least being able to use the output from one as input for another, will go a long way towards tying all this together. For example, Bill Nix made a very good comment that did not receive as much attention in the discussion as I think it deserved. He noted that it might make more sense to focus on going to more realistic time scales than to larger dimensional scales in certain types of atomistic calculations since this seems to be much more relevant to properties. There are probably many similar areas where input from one side of an interdisciplinary barrier would help to guide efforts on the
other side to more useful results.

Also, I think we all agreed that applying the principles of engineering science to biology is a very powerful direction to go.

I note that, although there were plenty of simulations, there were painfully few presentations of studies of real microstructures at the workshop. This, I think, is something that is also needed in the future.

If we get the chance to have another such workshop in the future, perhaps it would be interesting to have people explicitly present their visions for the future, as well as their current research directions.

Thanks again for organizing the workshop

with best regards,

-- Shef baker
_________________________________________________________________

Shefford P. Baker
Cornell University
Department of Materials Science and Engineering
Bard Hall 129
Ithaca, NY  14853

Tel:  +1 607 255 6679
Fax: +1 607 255 2365
_________________________________________________________________

-------------- Comments by Dr. Demitris Kouris---------------------------

Surface defects and thin-film growth

·       Defects (adatoms, steps, vacancies, etc) influence thin-film growth. The traditional thermodynamics of wetting does not provide a complete picture of the growth modes observed in a number of technologically important systems.
·       Recent experiments examining thin-film growth have demonstrated that local surface defect interactions can actually control the global film morphology. Surface defects corresponding to adatoms, vacancies, and steps together with misfit dislocations interact with one another affecting and often dominating kinetic processes.
·       Elasticity (continuum theories) can accommodate most of the problems involving long-range interactions. In most cases, however, the overall scale of the material system as well as the local geometry necessitate the use of atomistic approaches.
·       We can now determine the elastic interaction among individual defects with an acceptable level of accuracy (continuum, atomistics, quasi-continuum methods).

"Wish list" (things to be done)

·       Verify experimentally the significance of elastic effects; compare with the magnitude of the Schwoebel barrier.
·       Determine the magnitude of elastic interactions in complex systems; homogenization will not do!
·       Incorporate these effects in melecular dynamics and kinetic Monte Carlo simulations of growth.
·       Investigate the conditions under which nanomechanics can be used to tailor properties (electric, magnetic, optical) of solids at small length scales.
·       Understand the mechanics of electro-mechanical devices (sensors, motors, etc).
·       Develop a mechanics methodology for controlling adhesion properties.
·       Use nanoindentation as a tool to accurately characterize the mechanical behavior of solids/thin-films at the nanoscale; how can this information be used to determine the bulk solid performance?
·       Learn how to design materials from "the bottom up". What are the mechanics issues in the molecular assembly (attachment) of nanostructures (nanoparticles, motors, sensors, etc) and how do they effect the bulk device or structure they belong to?
 

          Professor Demitris Kouris
Mechanical and Aerospace Engineering
        Arizona State University
        Tempe, AZ 85287-6106
          tel: (480) 965-4977
---------------------------------------------------------------------------------------

------------------------Comments by Dr. Wendy. C. Crone--------------------------

Dear Kyung-Suk,

Thank you for inviting me to participate in the Nano and Micromechanics Workshop sponsored by NSF.  The experience was a very valuable one for me. I have included a few comments below.

Scientific Aspects:

The field of mechanics enables the application of basic scientific principles to engineering practice.  The workshop presentations made it clear that the application of mechanics to the nano and micro scales will allow us to take full advantage of the opportunities that devices and structures on these length scales provides.

One of the hurdles we currently face is the lack of a range of experimental techniques to draw upon.  Additional experimental techniques and instrumentation that conduct mechanical testing must be developed for the nano and micro scales so that essential physical quantities can be extracted.  The key challenge for experimentalists is to look beyond simply making our standard testing techniques smaller, we also need to think creatively about developing new test methods that lend themselves to the evaluation of material behavior at the nano and micro scales.  The focus must be to directly measure variables of real interest to study elastic and
plastic behavior at these scales.

The scientific issues we face must be attacked from experimental, analytical and numerical viewpoints simultaneously.  Currently, much of the funding structure of NSF makes this difficult.  The budget requirements of a truly interdisciplinary proposal which incorporates experimental, analytical and numerical work are too large for the programs which we target.

Educational Aspects:

>From the breadth of the research presented at the workshop, it is clear that fundamental understanding of mechanics is critical for all engineering majors.  The interdisciplinary nature of the research being conducted in mechanics also requires better mechanics backgrounds among affiliated disciplines.  In order to make significant changes in engineering education we must take on the task of convincing our collaborators and colleagues in other disciplines.  The value of mechanics can be shown in its application to fundamental research problems.  We must inform our colleagues of it usefulness.

Undergraduate and graduate degree programs in mechanics are also vital. The problem faced by degree programs in mechanics is that interest in this field has been small although the students that are attracted to mechanics form an elite group.  More high-caliber students must be recruited through a combined academic/industry effort.  Recruiting literature like "Engineering: Your Future" published by ASEE must be revised to include Mechanics as one of the engineering disciplines listed.   Industrial leaders must voice their need for students well trained in mechanics in order for high school and undergraduate students to consider the field in larger numbers.  The interdisciplinary nature of the field and the research being conducted is a fundamental point of interest that should be advertised.

To parallel our research, our students need to be educated in analytical, experimental, and numerical techniques.  All of our students certainly are required to master mechanics theory, but each student should also have some level of experience with experimental and numerical analysis techniques.

Yours,
Wendy Crone

--------------------------------------------------------------------------
Wendy C. Crone
Assistant Professor

Engineering Physics
University of Wisconsin
529 Engineering Research Building
1500 Engineering Drive
Madison, WI  53706

608-262-8384
[email protected]
http://www.engr.wisc.edu/ep/faculty/crone_wendy.html
--------------------------------------------------------------------------
"The most beautiful thing we can experience is the mysterious. It is the source of all true art and science."  Albert Einstein
--------------------------------------------------------------------------

------------------------Comments by Dr. K. J. Cho--------------------------

Dear Prof. Kim,

I have been thinking about replying to your mail on the NSF Workshop for quite a while. ...

Since I am new to this micro solid mechanics community, the NSF Workshop was a wonderful opportunity for me to learn about the major current research issues of this field. Personally, I was honered to chair the first session of the workshop, and I was very happy that it went well. Probably due to my background in solid state physics, I have noticed that the atomistic approaches (interatomic potentials and quantum simulations) are still not widely used, and I believe that atomistic simulation combined with new constitutive modeling for micromechanics would be a very fruitful direction to follow in the future. Based on this vision, my research group is currently studying thin film mechanics of metals and semiconductors.
...

KJ

----------------------------------------------------------
Kyeongjae (KJ)) Cho; Assistant Professor
Mechanics and Computation Division
Department of Mechanical Engineering
Stanford University
Stanford, CA 94305-4040.

Tel: 650-723-4354, Fax: 650-723-1778
e-mail: [email protected]
web: http://cdr.stanford.edu/html/ME/faculty/cho.html
----------------------------------------------------------

 

 

Opinions & Discussions