Design Optimization - Science topic

Dear Thuy,

In our last publication, we had the same problem with CCD and the alpha being generating negative values, so I inserted the levels in terms of alphas and it worked perfectly.

You can check our publication

Kind regards

Solution development based on proven technologies.

What are these proof-values: Reliability, ROI, TCO, ease of use.

How to check what's best: Monitoring, Predictive Measurement and Analytics, etc.

John Francis Miller Thank you!

This was my first idea, however I was concerned that such approach will interfere with calculations.

My new paper (it is free) can show it as a practical example (Acid Red 18 vs Acid Brown 14 and Acid Orange 7):

Techno-economical aspects of electrocoagulation optimization in three acid azo dyes’ removal comparison

I suggest you take a look on: https://www.mdpi.com/2226-4310/9/3/153

where aero-structural optimization is carried out using Particle Swarm Optimization.

See Chapter 7 in:

And also:

See Chapter 12 in:

Dear Murat Dörterler, Greetings!

This systematic approach gives the minimum volume of gear.

But you have to read in Russian.

You could define variables in the project and use them for all the simulations.

Thank you Mr. Rebbah but I think the problem is something else .

This is perfectly possible! At the moment we are developing it to optimse blade shapes of propellers.

Dear Phil,

The complexity of the refining business grows every day with new markets, new feedstocks and new regulations. Additionally, refineries are facing declining profit margins. I think, to sustain their profitability, refineries must leverage process simulation technology and capabilities to achieve best-in-class operational excellence. I think advance process simulation could be more helpful towards the inclining profit margin. Couple of areas where it can use effectively to reduce the market burden.

1. Heat exchanger maintenance and monitoring-thorough simulation of heat exchanger operations within the broader process simulation model. The heat exchanger design tool must also simulate all major heat exchanger types used in the refining industry. Furthermore, the solution should allow process engineers to easily develop and integrate their heat exchangers’ simulation as part of the refinery flowsheet without leaving their familiar process simulation environment.

2. Column operations troubleshooting-an integrated process simulator that accurately simulates the thermal and hydraulic behavior of the column unit to provide enough information to support column operations. With the correct process simulation software, users can accurately simulate thermo-hydraulic functioning of columns based on their construction and operating conditions. As a result, they can better understand the columns’ behavior and avoid operational mishaps. Simulating the operation of the column in the broader setting of the overall process enables users to identify root causes of the problems and determine the optimal point of operation for the overall process unit.

3. Integrated refining and gas plant analysis-Refineries need a solution that meticulously simulates the entire gas plant including acid gas treatment units, sulfur recovery, tail gas units and flare systems together with the mainstream refining process units, such as distillation units and reactor units. Advanced simulation technology would provide the refiners enough confidence to push the levels of sour crudes closer to the limit the refinery can process while meeting regulations. Feed flexibility, capacity creep and operating expenditure optimization, enabled using integrated refining and gas plant process simulation, can save refiners millions each year in operating margins while ensuring maximum reliability and plant uptime. In addition, the rigorous simulation of the gas plant operation offers refineries visibility and the ability to better document their emission levels. This capability is valuable for boosting their profit margins.

4. Planning model update for refineries-The ideal option is to give refinery process engineers the ability to maintain the planning models with the help of advanced process simulation software that can offer a streamlined workflow to update the planning models, enabling frequent updates when the models become out of sync with the operating range of the refinery.

5. Refinery-wide process analysis-With an advanced integrated solution for process simulation and refinery planning, refineries can develop a refinery-wide process model out of their refinery-wide planning model in a relatively short period of time. The accuracy of the simulation model can be enhanced by selectively incorporating rigorous models of reactor units to the refinery-wide flowsheet.

Ashish

Optimum result from factorial design gives the good results, but takes more computation time.

I recall the book

Wendland, H.: Scattered Data Approximation. Cambridge Monographs on Applied and Computational Mathematics, Cambridge, UK (2005).

A PhD student of mine used it thoroughly in her PhD project on simulation-based optimization in industry.

Check it out!

Hi Nazim Nariman

I am studying about the above topic!

I prepared a complete tutorial about optimizing artificial neural networks using genetic algorithm with Python implementation. It is titled "Artificial Neural Networks Optimization using Genetic Algorithm with Python" It is available here:

https://www.kdnuggets.com/2019/03/artificial-neural-networks-optimization-genetic-algorithm-python.html

https://towardsdatascience.com/feature-reduction-using-genetic-algorithm-with-python-403a5f4ef0c1

You can read more and more in my 2018 book that covers GA in one of its chapters. The book is cited as “Ahmed Fawzy Gad ‘Practical Computer Vision Applications Using Deep Learning with CNNs’. Dec. 2018, Apress, 978–1–4842–4167–7” which is available here at Springer https://www.springer.com/us/book/9781484241660

Also here at Amazon: https://www.amazon.com/Practical-Computer-Vision-Applications-Learning/dp/1484241665/

Samuel Ayinde @ Shape optimization with abaqus is very comman by using tosca. Tosca helps to improve existing design/shape for more reliable and durable components based on stress reduction through shape optimization. It also used for topology optimization for additive manufactured product. Here i am attaching abaqus/tosca manual: http://meet.cadcam-group.eu/pdf/SIMULIAseminar_April20/Georgi_Chakmakov_Tosca_Optimization_techniques.pdfThis attachment shows structural component design, shape and thickness optimization, which finally tells that there is weight reduction. This weight reduction means less material requirement, then the overall cost also decreases.Please also find the attachment tutorial for the same: https://www.4realsim.com/tosca/The relevant paper for optimization in the abaqus environment using TOSCA: http://www.simulia.com/download/pdf2009/Furbatto_SCC2009.pdf Tutorial for design optimization: https://www.youtube.com/watch?v=R5BFd5c4pfE

Tutorial for topology optimization: https://www.youtube.com/watch?v=vKmFfHPkWtw

Comsol is flexible in customizing the geometrical parameters.

Please look at few reference literature realted to your question using COMSOL software.

http://www6.cityu.edu.hk/appkchu/Publications/2015/15.88.pdf

Good work if results true obteind

Hi Samuel

Are you familiar with these publications:

Topology optimization with a level set approach in Matlab [1]

Topology optimization with a density-based approach in Matlab [2] 

I think it provides, what you are looking for.

[1] https://link.springer.com/content/pdf/10.1007%2Fs00158-009-0430-0.pdf

[2] 

Few more articles

'Greater Chinese'global production networks in the Middle East: the rise of the Jordanian garment industry

S Azmeh, K Nadvi - Development and Change, 2013 - Wiley Online Library

The expansion of ‘Greater Chinese’ capital from mainland China, Hong Kong and Taiwan into other parts of the developing world is increasingly noted. It is especially prominent in sub-Saharan Africa where Greater Chinese investments, firms and workers are found across a wide range of activities, from the extractive commodity sectors, to infrastructure projects, agriculture and manufacturing. One region where Greater Chinese investment is less well studied is the Middle East. This article focuses on the case of Jordan. Jordan has rapidly emerged as an important supplier of apparel to the United States, a consequence of a distinct preferential trade agreement. The article charts the ways in which this preferential trade agreement has stimulated the shifts of Greater Chinese garment manufacturers to Jordan. Using a global production networks (GPN) framework, and drawing on primary and secondary evidence, it assesses the dynamics behind Greater Chinese investments into Jordan; it also explores the ways in which Greater Chinese garment producers operating in Jordan organize their supply chains and are linked into the global garments GPNs. Finally, it considers the relationship between such capital flows and the influx of Asian migrant workers into the Jordanian export garment sector.

The Cases: Korean Vehicle Manufacturers in Europe (The 1990s)

JH Hyun - Korean Automotive Foreign Direct Investment in …, 2003 - Springer

This chapter attempts to trace the activities of Korean VMs in Europe. It is suggested that the internationalisation process of a company should be understood at the corporate level; the company’s specific history and changes in the competitive and regulatory environment (Nilsson et al. 1996). This study, therefore, selected two Korean companies in Europe as particular cases. These companies are HMC and DMC. In 1997, both companies comprised over 70 per cent of the country’s total in terms of production, sales and exports (Table 6.1). KMC is considered as a part of HMC based on the fact that Hyundai merged with KMC in October 1998.

A manufacturing network for generating added value from a geographical distance for the next generation

K Yukawa, T Kawakami - Journal of Machine Engineering, 2011 - yadda.icm.edu.pl

this study, we suggested a manufacturing network for generating added value from a geographical distance in terms of value creation from various management resources. Until today, most researchers focused on the Real Concentration of Production Base, which provides certain manufacturing benefits in these domains. However, "the Virtual Concentration of Production Bases" is realized by overcoming large physical distances and time differences between production bases, and creates greater added value for products. We attempted the simulation of the manufacturing network of creating the added value after having shown the new framework of network analysis in the manufacturing system.

Geography of multinational corporations and functional specialization in Chinese cities

C He, X Xiao - Symphonya, 2011 - search.proquest.com

Function evaluations are one of the most important criteria for comparison along with statistical analysis such as standard deviation, Friedmann's Test, Wilcoxon(due to the stochastic nature of metaheuristic), minimum fitness as well as average fitness. Function evaluation is important since several metaheurisitics have 2 or more stages in their optimization process (for example TLBO has teacher and student phase, GWO has only phase (searching and hunting), ABC also has two loops, Jaya has one , BFO has 5 etc). For all of them number of iterations may be same but that does not reflect the true computational requirement of the algorithm in terms of how much effort was required by the algorithm. Therefore, by using functional evaluations, a more fruitful comparison can be drawn since we are aiming for a common ground, basically the number of times the fitness function was called.

Thanks a lot Prof Mohamed-Mourad Lafifi for your time and suggestions . This is a great addition to my knowledge bank.

Dear Vahid, the direction of the displacement is correct and it follows my velocity direction, I got the problem which is referring to the wrong connection of my rigid region coupling and related DOFs.

Thanks for your attention.

Thanks for the quick answer!

Maybe I should add some additional explanation of my application, as the question was a bit inaccurate.

I have a pointset with 2000-2050 points, they are ordered in x direction from right to left over a distance of 100 mm with an average distance of 0.05 mm between two neighbouring points (the points are ordered like pearls on a string).

They come from a laser pofile scanner (works with triangulation), which measures a straight line of a mechanical part. With the help of the pointset I want to identify the manufacturing precision of the part. For this purpose I want to measure the straightness of the line (standardized term in tolerancing), which could be defined as the "height of the line profile's bounding box with minimum height". Therefore I calculate the convex hull and measure it's 'minimal widht'.

A typical value of my straightness is 0.2 mm, so the bounding box has height 0.2 and length 100.

My problem now is, that the laser profile scanner has some uncertainty, which is some kind of random deviation of the point's position. This deviation mainly affects my straightness value, if points are deviating "away form the pointset", as the number of convex-hull-points is very small (only 10-20) compared to the number of all points and the requirement for a point to have an impact on the straightness is that it has to be part of the convex hull.

The problem with statistical estimation is, that i don't know how to estimate a value from a series of measurement, where all measurements are overestimations.

The idea with 6 standard deviations is very intersting. I think i first have to have a look at the range of my measurement error and subract it two times from my straightness value.

Hi Michael,

That picture is quite interesting - thank you for uploading it.

OK my original advice about switching to 2-D analysis is inappropriate here because the pores in your sample are irregularly shaped, i.e. there is no symmetry at all, let alone the particularly powerful type of symmetry that allows you to convert your 3-D simulations to 2-D simulations (axisymmetry). So ignore that if you haven't already.

With respect to modifying the interior topology, I have a further question. Are there any particular constraints? For instance, do you need the same volume fraction of pores in the polymer sample as in the original biological sample? Or the same surface area to volume ratio, number of vertical strands, strand diameter or similar?

My thought was that these constraints can be manipulated into mathematical expressions in the input file (Python) that ABAQUS can pick up, generate the geometry, mesh and input files automatically for and then optimise a particular result (your design variable), e.g. the reaction force at the top or bottom surface. Or did you have a different idea in mind?

Apologies - I am playing catch up here.

Milan.

Yes. In fact there are some tutorials on YouTube. If R2014a features optimization toolbox, you can easily optimize your functions with its help.

Dear Jerry,

Please see answer of similar question trough Researchgate here is link:

How can I optomize an analytical procedure by response surface ...

https://www.researchgate.net/.../How_can_I_optomize_an_analytic...

Also i suggest to you attached files in topics.

Best regards

Hi, what do you mean by "issues in calling multi-variable function" ? I suppose you have coded this function in MATLAB or in another language. Is the problem related to the link between your function and the PSO code ?

You may follow Mahamad Nabab Alam and read his paper and algorithm as in

Codes in MATLAB for Particle Swarm Optimization

Thank you

Dear Katharine,

I am very grateful for your collaboration. Thank you very much!.

Kind regards,

Pablo.

Taguchi L27? Does that mean you are trying to optimise a response with respect to 13 independent factors from 27 observations?

Hoping this helps you. Good luck.

I can confirm that for a minimum-time transfer the thrust magnitude is constant and maximal. For a minimum-fuel transfer, the thrust magnitude alternates from maximal-thrust and zero-thrust.

Back to your question, I can only speak for electric propulsion, which is usually referred as low-thrust propulsion. In general, the thrust of an electric propulsion thruster can be throttled. Depending on the thruster type (arcjets, RITs, HET etc.), the throttle characteristic is very specific. The operation region varies from type to type. The power and/or the specific impulse shift from the initial operation point to one direction or the other. Every thruster type has its very own specific throttle characteristic. The challenge is modeling this characteristic in your optimization problem to find your optimal steering law.

Another aspect is the benefit-cost ratio. A thrust variation might save a fraction of fuel mass, but the downside is a more complex control of the propulsion system and a much more complex qualification process of such a thrust profile. Sometimes different operation points are given by the suppliers, see link.

I am answering this question with respect to linear optimization.

Multi objective functions will take more than one objective functions. Let us say we want to maximize contribution (to profit) and maximize volume (in no of units) for a multiple product company. So it is necessary to use something like giving goal so that I can know how much weights we want to give each goal. So in a multi- objective framework it is necessary to assign right weights to the goals.

The slope of some of the variables, I understand that it is the coefficients will be different.

Let us say objective function is = w1(f1(X) + w2(f2(X)  

= w1(C11X1+ C12 X2+...+C1nXn) + w2 (C21X1+....+C2nXn) 

Given this , only requirement is how to assign the w1 and w2 and not C vector for 

Here we must understand that 

1) Should we solve for one goal or more than one goal . Optimal Solution for profit maximization will not be same for solution for utilization  maximization  

2) Optimal Solution for profit maximization will be different from that of cost minimization.

In my opinion finding the weights (W1 and W2) are more important that finding the

3) The coefficients are generally provided are  parameters. 

I hope it helps 

Goutam Dutta

Thank you very much for sharing,

Good luck

Dear Prof., we have worked on triple objective engineering design problems in the past. I am including some of our works, hopefully this helps you:

Ganesan, T., Vasant, P. and Elamvazuthi, I., 2013. Hybrid neuro-swarm optimization approach for design of distributed generation power systems. Neural Computing and Applications, 23(1), pp.105-117.

Ganesan, T., Vasant, P. and Elamvazuthi, I., 2013. Normal-boundary intersection based parametric multi-objective optimization of green sand mould system. Journal of Manufacturing Systems, 32(1), pp.197-205.

Ganesan, T., Elamvazuthi, I. and Vasant, P., 2015. Multiobjective design optimization of a nano-CMOS voltage-controlled oscillator using game theoretic-differential evolution. Applied Soft Computing, 32, pp.293-299.

T Ganesan, I Elamvazuthi, 2016,A Multiobjective Approach for Resilience-based Plant Design Optimization, Quality Engineering.

Thank you.

In the aerospace industry, where safety is paramount, both material properties and high metal removal rates are important; one does not choose one over the other.

In the 1980s Rolls-Royce had very low removal rates on compressor and turbine disks. The view was to extend tool life and hence lower tool cost. However, they were very uncompetitive compared to GE and P&W who used much faster cutting techniques and throw away tools; the tool cost was insignificant compared with the product cost. The view at that time was that fast cutting would decimate material properties. However, when they did the experiments and had both properties and material removal rates as constraints, they could meet both requirements.

The research led to new and better cutting materials which put them in the lead (for a while anyway). This along with other innovations allowed them to go from a poor third to first or second (depending on how you measure it) gas turbine manufacturer in the world.

You can define a Hard Point in area

The critical part in CTAB is the NaCL concentration and amount of polysaccharides in the plant material. As long as you can balance it well, you can get good amounts. I would be tempted to use Liq Nitrogen grinding to break open the plant cells more effectively.

So in short, make sure the NaCl concentrations are right, perform two or more extractions in CTAB depending on ur polysaccharide content.

Though isoprop is better for precipitation, I would suggest a couple of rounds of ethanl washes to remove impurities.

It all depends on ur downstream application.

Checked the document. Could not find any reference to BERT. I appreciate your efffort oif helping me.

The Optimization by the AG according to unknown to the objective function to be optimized

best regards

In general I agree with the answer of Claes Richard. This is the right procedure.

What you always need is a communication channel between the analysis solver (FEA) and the optimizer (Matlab). Because the output of FEA (stresses, displacements, etc) is the input for the optimizer and the output of the optimizer (new design variables) is the input for FEA and this goes on in cycles until convergence. This communication can be done either through memory (when you write code for both FEA and optimization), DLL communication or (the most simple solution) text files.

Given that Matlab can read/write to text files without any problem, you need a FEA analysis tool that can be called as exe from Matlab with the right arguments (analysis options, location of input file, etc), can read its input text file (prepared from matlab optimizer), analyze the model and produce an output text file (to be read by matlab optimizer for optimization constraint checking).

Any built-in matlab optimizer can be used (GA, SQP, etc) with any such analysis tool. It is rather easy to do the coupling yourself and this coupling can provide very impressive results!

The area or the volume calculated using isogeometric analysis (NURBS-based discretization) is accurate. The calculation follows the standard numerical integration scheme. You firstly discretize the domain or region using finite element method or IGA, Then you integrate the area or volume of  each element. After that , you summarize them and you can get the volume or area. The integration scheme follows a similar way that you calculate  the local or global stiffness matrix. You may find the discrete shape gradient of the volume/aera with respect to the design control points in one of my papers about the isogeoemtric shape optimization. good luck.

I agree with Paul. and add to him, There is a device noise and circuit noise. In order to verify the performance parameters of a device you have to stick strictly to the given instructions in the manufacturer data sheet for measuring such parameters specially the noise performance.

The noise given in the data sheet is the root of the spectral noise power density in nV/ square root of Hz.  So if you measure the total noise you have to integrate the spectral noise density over the amplifier bandwidth.

You have to take into consideration the noise generated by the external input resistances or input sources. As Paul; said you have to choose metal film resistors which is less noise that the carbon film resistor. Also, the external resistance at the amplifier input must be as small as possible. Also, it may be very helpful to restrict the bandwidth of the amplifier to just the required frequency range.

Concerning the ripples you must suppress then by good regulation and filtering.

One reaming  note, what do you mean by the stability? is it the frequency stability?

or what? or noise? 

wish you success

Placket-Burman is not suitable for optimization. The purpose of using Placket-Burman is to determine the most significant factors (Screen design). In general, Placket-Burman is not recommended because interactions are partially confounded or "aliased" with main effects.  Composite design is mainly used for optimization purposes with no confounding. Since you have 8 factors, CCD requires around 282 runs (Full design) which is very expensive and time consuming. Therefore, I recommend you to use first Placket-Burman to find few critical factors for the output responses, and then use CCD to optimize these critical factors. Please do NOT use face centered cubic design which is one of response surface designs because this method has multicollinearity if the problem has more than 5 factors. 

You can calculate it using the Hollomon equations

Y=K*eps^n

in which K and n depends on the material and eps is the deformation you have.

Franco

Yes they are not usually that critical. 

The work is described in my two recent papers. You can download them from my researchgate page.

InsyAllah I will explain in detail soon.

Thank you all for your input. One of the reference that I found help is:

Nunes, L. O.; Martins, W. A.; Lima, M. V. S.; Biscainho, L. W. P.; Lee, B.; Said, A.; Schafer, R. W., "Discriminability Measure for Microphone Array Source Localization," Acoustic Signal Enhancement; Proceedings of IWAENC 2012; International Workshop on , vol., no., pp.1,4, 4-6 Sept. 2012

Dear sir,

Thanks for your answer

My requirement is slightly different from what I mentioned earlier. I am looking for a method/algorithm/option to obtain "axis aligned bounding box co ordinate values "

Either, but check wich options of endonucleases will cut/not cut your SNPs. You might find several options including isoschizomers that are cheaper or work better. Also worthed is to try the new enzymes that do the job in 15minutes.

The answer on the question is depended on kind of material for pyrolysis. Proposing it is usually waste tire, the answer is next depended on the size of that – is it shredded for chips of around 50mm size, or pieces of 200-250mm size, or even in kind of whole tires? To be short, the first approach is complicated with a tire shredding but it is most sure for continue processing with all of safety, health and environment standards to be observed, being carried out usually in a rotary kiln or auger pyrolysis reactor presented in one of the first indicative US patents such as 563580 (1997). Since 2007 the reactor auger design has been developed for double or even triple-set unit as you can see in my papers where use of steam that is co-generating and superheating with pyrolysis reactor heating, it is another question for consideration.

In the case of Y1, if Prob>F is lower than 0.05, the lack-of-fit is not significant. Contrary to what you say about Y2 where Prob>F is higher than 0.05. Anyway, if variables Y1 and Y2 are not significant but the interaction is significant, staticians say that the final model must have both the pure variables and the interaction in the empirical equation to be real. Let your software propose the otimization design points. I hope this comment will be helpful.

You can consult the book "Applied mathematical Solving Problems with MATLAB" of the authors: Dingyü Xue and YangQuan Chen, where there is a example problam that answers your question.

Please have a look at the following articles: You could search for the authors (Barros and Leps) and you will find more articles. 

de Medeiros, G. F., & Kripka, M. (2014). Optimization of reinforced concrete columns according to different environmental impact assessment parameters. Engineering Structures, 59, 185-194. doi: 10.1016/j.engstruct.2013.10.045

Barros, M. H. F. M., Martins, R. A. F., & Barros, A. F. M. (2005). Cost optimization of singly and doubly reinforced concrete beams with EC2-2001. Structural and Multidisciplinary Optimization, 30(3), 236-242. doi: 10.1007/s00158-005-0516-2

Lepš, M., & Šejnoha, M. (2003). New approach to optimization of reinforced concrete beams. Computers & Structures, 81(18-19), 1957-1966. doi: 10.1016/s0045-7949(03)00215-3

EL DEBS, A. L. H., NETO, A. C., CHAVES, I. A., SQUARCIO, R. M. F., & LIRA, S. A. Optimization of cross section of reinforced concrete beam using experimental design.

physiological parameters are also imp. as the nutrients of production medium. Temp., rpm, pH etc are those which can be optimized first as one factor at a time den u use PBD for other sources like c source and nitro sources and also mineral salts.

yes u can use each sources separately but it is recommended to use above 4 sources for each section as PBD rule says so.

All d best...

Thanks lot mem

You can try Unscrabmler software. The CAMO company offers you 30 days free trial of software. You may find the required details in the help column.

with best wishes

Optimization is a very specialized field and can also be a complex area. As I was teach from the beginning, one main point is that the choice of the adequate algorithm depends on the formulation, as well as how to formulate it depend on the algorithm to use.

With this, I will try to start answer this question with my perspective and hope others, more specialized than me, may complement this answer or give others.

In deterministic optimization the input data for the given problem is known accurately. It is considered e.g. in many structural optimization problems.

There are a plenty of problems where the input data for the given problem is NOT known accurately. Then, the optimization under uncertainty, or stochastic optimization, is chosen in a way that the uncertainty be considered under assumption of a given (assumed known) probability distributions. E.g. it is the case where the human behavior has influence in the performance being optimized like for e.g. in economic applications.

Exists an intermediate special type of problem where parameters are known but only within certain bounds. For theses cases Robust optimization techniques are usually more adequate. E.g. cases where a structure is sometimes loaded differently in different scenarios, which do not occur simultaneously. In robust optimization, is like one is looking for a compromise between different optimums found in each scenario without calculating necessarily these different optimums.

It is not unusual that one gets a kind of problems that are computationally very very hard. Then the time spent in computations (e.g. line search techniques) is not adequate and there is some evidence that for these problems, heuristic search techniques (e.g. randomness is used to find better solutions) are better methods. Examples of heuristic optimization methods are simulated annealing, genetic algorithm and evolutionary algorithms.

If the examples mentioned are too generic, particular applications of specific interest can be easily found (with some spent of time) in the research literature.

actually the concept of the study should be bit more practical and not just solving problems rather than sitting and solving complicated numerical .... lot of software are available now a days .... only thing required is the proper interface of the students with the concept of the topic of designs and QbD.

You can use this two types of designs. They are very powerful and useful. But not flexible if you want to define your model under study and if you have a limited number of runs to perform.

So I believe that you should construct the design for your particular situation using optimal design of experiments. You can find these functions in the JMP Software (free trial). You will need your model under study, the number of runs, among other things.

Now, to use the methodology in a correct way, you need to state what is your study for. These are some possibilities:

- Construct a design that allows you to have good estimates of your parameters. Using D-optimal designs

- Construct a design that allows you to have good prediction power. Using I-optimal designs

- Construct a design to make some screening to your factors (quadratic, interactions, main). Definitive Screening Designs See http://blogs.sas.com/content/jmp/2012/01/30/introducing-definitive-screening-designs/

Karnaugh Maps (K-Maps) and Quine-Mc Clusky algorithms remain the best ways to reduce the digital designs in their minimum forms ( Sum of Product ) or (Product of Sum). However, I don't like the use of don't care terms as being arbitrarily logic 1 or logic 0. Don't care terms create spurious outputs if they are clubbed with 1s in KMAP minimization. Different computer programs have been developed to do the logic reduction for KMAP or Quine Mc Clusky.