Assessing Novice CAD Model Creation and Alteration

9 Assessing Novice CAD Model Creation and Alteration Xiaobo Peng1, Prentiss McGary2, Michael Johnson3, Bugrahan Yalvac4 and Elif Ozturk5 Prairie View A&M University, xipeng@pvamu.edu Prairie View A&M University, prentissmcgary@yahoo.com 3 Texas A&M University, johnson@entc.tamu.edu 4 Texas A&M University, yalvac@tamu.edu 5 Texas A&M University, alpe44@neo.tamu.edu 1 2 ABSTRACT To allow for the efficient reuse of existing CAD models, design intent has to be clearly conveyed in the original design. A modeling and alteration exercise was implemented in a freshman class using NX to assess the CAD modeling procedures of novice CAD users. To analyze the effects of CAD model attributes on alteration time and procedure, the attributes of both the original and altered models were analyzed. A survey was conducted which assessed students’ perception of the original models. The analysis data and survey results are presented in this paper. The results show the effects of different incentives on modeling and how the feature selection and organization correlate to proper design intent. This work provided base data for the study to form preferred CAD modeling procedures. Keywords: CAD, design intent, attributes, NX. DOI: 10.3722/cadaps.2012.PACE.9-19 1 INTRODUCTION Today the digital computational tools, including Computer Aided Design (CAD)/Computer Aided Manufacturing (CAM)/Computer Aided Engineering (CAE) tools, have become ubiquitous in the design and manufacturing industries, like GM, Ford, Boeing, and 3M etc. [10] The use of these tools can reduce product cycle time and increase quality. The product design processes in the industry have proceeded more and more as a “virtual product” generation process enabled by the CAD tools [4]. Today’s industries demand that universities graduate engineers who are prepared with the skills of using appropriate modern CAD/CAM/CAE tools [5]. A student who is not knowledgeable in using CAD/CAM/CAE tools will be placed in a distinct disadvantage after graduation in this highly competitive environment [10]. The training of these tools has been considered as mandatory for engineering students. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 10 Although the advances in CAD tools have changed the practice of professional engineering, the curriculum for engineering education has not kept pace in adapting such technologies [5]. According to a survey conducted by Ye et al. [17] on how industries evaluate the current CAD education in colleges and universities, 74 percent of the participants from the industries indicated that current CAD education is inadequate. Most current CAD instructions are focused on teaching declarative knowledge – the key strokes and button picks required to perform certain tasks in specific software platforms. Very limited attentions have been paid to teach students how to create a good model which can be easily understood, altered, and reused by others [8], [9]. Many of the new products are based on the reuse or alteration of the previous designs [12]. CAD allows the easy reuse of existing components and assemblies. To allow for the efficient reuse of existing CAD models, design intent has to be clearly conveyed in the original design [8]. To leverage the power of modern CAD tools and create models that are easily understood and altered, students must learn to properly convey the design intent of a given model [15]. To address the deficit of the CAD education, the authors have worked on the creation of empirically-derived prescriptions and instructional activities to improve students’ CAD skills. This paper presents the initial work of implementing a CAD modeling and alteration exercise adopted from Johnson’s previous work [8],[9] in a freshman mechanical engineering class. The goal of this work is to assess the CAD modeling procedures of novice CAD users and analyze how the feature selection and organization correlate to proper design intent. Ultimately our goal is to form preferred CAD modeling procedures. The rest of the paper is organized as follows. In Section 2, related works are reviewed. Section 3 presents details on methodology. Analyses and results are discussed in Section 4. Discussion and limitations are presented in Section 5. Conclusions and future works are described in Section 6. 2 RELATED WORK Literature has suggested that a skilled CAD user should possess two distinctive skills, namely declarative knowledge and procedural knowledge (also referred as strategic knowledge) [2], [3], [6], [11], [14]. Declarative knowledge is the knowledge of the commands which can be used on a particular CAD system. It is the knowledge of knowing that or knowing what [11], [14]. The procedural knowledge is the knowledge of strategies or alternative methods, which can be used to carry out a task in CAD. It is referred as the knowledge of knowing how. The design intent would fall under the category of strategic knowledge [3], [9]. Design intent encompasses all of the major decisions related to CAD modeling: feature selection, order, and organization [13]. Design intent can broadly be defined as the purpose or rationale behind the choice of a given design object [7]. It includes the constraints and parameters that define the features and relationships between features [9], [14]. In the work conducted by Lang et al. [11], the procedural knowledge was extracted from keystrokes analysis. They concluded that the procedural knowledge is transferable between CAD systems. Chester [3] presented an instructional experiment with novice CAD learners where the specific procedural command training and spatial ability training were introduced to enhance students’ CAD strategic knowledge. Rynne and Gaughran [14] presented a theoretical frame work to define the cognitive aspects of the modeling tasks. Their findings indicated that students can better convey design intent if they have the ability to visualize and deconstruct objects and to cognitively assemble them. In turn the users become more efficient in using the CAD systems. Hamade et al. [6] collected performance time data through CAD exercises using Pro/Engineer. Declarative and procedural learning curves were generated respectively from the data. The results showed that the Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 11 procedural and declarative components of CAD learning are largely cognitive. Wu [16] carried out a similar work to study the learning models of industrial design students. His results backed up the findings of Hamade et al. Recently Alducin-Quintero et al. [1] used annotations in CAD models to improve design intent communication. Their work assessed the impact of annotations in the engineering change process. The experimental results showed that the time to perform engineering changes in existing models was reduced by 10-20% when providing annotations in the CAD models. All the aforementioned researches suggested that strategic knowledge or design intent is critical to the efficient use of CAD systems but is not generally highlighted in CAD education and training. The strategic knowledge is even more important for novice learners of CAD because applying design intent early in the CAD modeling may prevent later model failure. In addition, it is possible to embed strategic knowledge instruction into command teaching without affecting command knowledge [3]. However, most of the researches were focused on assessing the impact of the strategic knowledge in CAD performance. There is little comprehensive empirical work in the area of relating model attributes to design intent. While some CAD modeling prescriptions to improve the communication of design intent have been proposed by Hamade et al. [6] and Rynne and Gaughran [14], they are limited and usually not backed up by empirical evidence. Most recently, Johnson and Diwakaran [8], [9] have performed empirical experiments to evaluate the modeling procedures and attributes that facilitate model understanding and alteration. The experiments were carried out in junior and senior level CAD related courses using Pro/Engineering and SolidWorks. This paper presents the preliminary findings of the CAD modeling and alteration exercise implemented in a freshman level course using NX. The ultimate goal of all these works is to form preferred CAD modeling procedures to best convey design intent in cross-platform CAD systems. 3 METHODOLOGY To assess the CAD modeling and alteration processes of novice users, an exercise was carried out in the freshman CAD course in Mechanical Engineering. It is a 3 hours laboratory class where students learn engineering graphics and 3D modeling based on NX. The modeling and alteration exercise was implemented in two weeks towards the end of the semester when the students had achieved significant familiarity and competence with the software. During the exercise, the students were divided into two groups based on their performance in the class so that two groups had similar CAD skill level distribution. The exercise was conducted in two phases. Phase one is the modeling phase. Phase two is the alteration phase. In phase one, both groups were asked to build the original model as shown in Fig. 1 with different goals. The goal of group one (denoted as speed group) was to model the part as quickly as possible. The group two (denoted as alteration group) was told that their goal was to design the part so that it could be easily altered by another member of the class next week. The students were given 100 minutes to complete their models. Their time of completion was recorded after the instructors checked the accuracy of the models. Students were given extra credits to participate in the exercise based on their performance. Five, three, or two points were given to the students in top third, second third, and bottom third of each group respectively. The students were not aware of the altered design when they were creating the original models. In phase two, the same two groups of students were asked to conduct the alteration exercise. The models of the students who finished in the top third of each incentive group in phase one were distributed to the other incentive group to be altered; three students attempted to alter each of the models. The students were again given 100 minutes to alter the original models to the model shown in Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 12 Fig. 2. The students were not made aware of which incentive group provided the model they were altering. Their time of completion was recorded again. After the alteration phase, the students were asked to complete an anonymous survey to assess the designs that they had altered. The survey asked students to rate the design using three metrics in seven point scale (1 as the lowest rate; 7 as the highest rate). Specifically, they were asked to rate the intuitiveness of organization, feature order of the model, and overall rating to the model. The survey also asked the students to rate how helpful certain modeling attributes would be to their alteration of the original model. This was again done on a seven point scale (1 – would make the model much worse; 7 – would be very helpful). The attributes assessed included: the naming of features; the use of more complex features; the use of simpler features; the use of patterns and mathematical relations; the use of copy and mirror features; and the use of datum geometry for referencing features. The surveys were collected based on the incentive groups. To analyze the effects of CAD model attributes on alteration time and procedure, the attributes of both the original and altered models were evaluated. The attributes were based on those described by Rynne and Gaughran [14] and Johnson and Diwakaran [9]. They were slightly modified to reflect aspects unique to NX based on the expertise of the authors. These attributes and quantities are listed and described in Tab. 1. The attributes of the original and altered models were compared for each incentive group. The correlations between the attributes of the original models were assessed. The survey results were also analyzed. The detailed results are presented in the next section. Fig. 1: Drawing of the original design. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 13 Fig. 2: Drawing of the altered design. 4 RESULTS The results for the completion time are listed in Tab. 2. In the modeling phase, there were 8 students in the speed group and 9 students in the alteration group. Four students in each group completed the original modeling exercise in 100 minutes. The completion rates were 50% for speed group and 44.4% for alteration group. The average completion time for speed group was 69.6 minutes (standard deviation was 20.3 minutes). The average completion time for alteration group was 71.5 minutes (standard deviation was 17.3 minutes). The alteration group had an average completion time that was 1.9 minutes greater than speed group, but the difference is not statistically significant (t=0.14, p=0,446). In alteration phase, the speed group (group one) worked on the models originally designed by alteration group (group two). The alteration group worked on the models originally designed by speed group. Only 1 out of 8 students in the alteration group completed the alteration task. 2 out of 6 students in the speed group completed the alteration task. The completion rates were 33.3% for speed group and 12.5% for alteration group. The average completion time of the alteration group was 23.7 minutes longer than that of the speed group. The difference is statistically significant at the p<0.1 level (t= 4.67, p=0.067). The total time (modeling and alteration) for the models originally designed by the alteration group was 134.8 minutes, which was 21.8 minutes less than the time for the models originally designed by the speed group (156.6 minutes). The preliminary results agreed with the predicted trend that the incentive has an effect on the completion time of CAD modeling. The completion time and rate show that the models originally created by the alteration group were easier to alter than the models created by the speed group. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 14 Attribute Correct Initial Sketch Plane Correct Model Origin Correct Base Feature Correct Part Orientation Correct Feature Sequence Number of Features Use of Reference Geometry Simple sketch and feature geometry Incorrect Feature Terminations Number of Pattern Features Number of Mirror Features Number of New Features Number Of Features Deleted Percentage of Feature Retained Number of Features Unchanged Number Of Features Changed Description Denotes whether the sketch for main block feature is placed on the proper datum in the model Center of main block feature located at global origin Main block as first (non-datum) feature Proper orientation of part in the model Should begin with main feature and end with ancillary features (e.g. chamfers and rounds) The total number of features. Sketches are not counted as additional features; pattern features include the pattern, the original feature, any additional required geometry; mirrors are counted as a single feature; all datum features (outside default planes and coordinate system) are included All datum features (outside default planes and coordinate system) Average number of sketch segments per extrusion or revolve; rounds and chamfers per feature Number of features that do not have correct feature terminations (e.g., through holes not defined as such) Measure Binary: 0 – no. Binary: 0 – no. Binary: 0 – no. Binary: 0 – no. Binary: 0 – no. 1 – yes; 1 – yes; 1 – yes; 1 – yes; 1 – yes; Whole number Whole number Real Number Whole number All pattern features Whole number Includes both solid and sketched mirror features Whole number The number of new features added to alter the model to the final design The number of features in the parent model that were deleted during alteration The percentage of features from the parent model that are retained with or without changes made The number of features that have been carried over to the altered model as is from the parent model The number of features that have been modified in the altered model Whole number Whole number Real Number Whole Number Whole Number Tab. 1: Descriptions of assessed model attributes. The attributes of the original models were then evaluated. The data were analyzed to examine any statistically significant difference between two groups as shown in Tab. 3. All the students in both groups used the right sketch plane, base feature, and orientation. However, none of the students used the right origin and the correct feature sequence. The alteration group used less number of features. This contradicts to the results in previous work [9]. The alteration group used more reference geometry features. It would be expected as using the reference features will cause more modeling time and help the models easier to be altered. The speed group used more segments per feature and more incorrect feature terminations than the alteration groups. This would be expected because one can create a model faster without defining correct feature termination, but it will make the model harder to change. These differences are in agreement with the previous work. However, all these differences are not statistically significant as seen in Tab. 3. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 15 Speed Alteration t Total Students Participating in Phase one 8 9 - Significance (1-tailed) - Number Completing Modeling Incentivized for 4 4 - - Average Completion Time (minutes) 69.6 71.5 0.14 0.446 Standard Deviation 20.3 17.3 - - 47.8 55.7 - - Alteration Speed t Significance (1-tailed) Total Students Participating in Phase Two 8 6 - - Number Completing Alteration 1 2 - - 12.5% 33.3% - - 87 63.3 0.067 N/A 4.13 4.67 - 87 60.4 - - Minimum Group Completion Percentage Average Alteration Time (minutes) Standard Deviation Minimum Alteration Time - Tab. 2: Completion time for speed group and alteration group. The correlations between the various model attributes for the original models were analyzed as shown in Tab. 4. Most of the correlations were not statistically significant at p<0.1 level. It is worth to mention that the original modeling time was positively correlated with the number of features and negatively correlated with the number of complex features (i.e. higher average number of segments per feature). This is in agreement with the results of Hamade et al. [6] and Johnson and Diwakaran [9]. The modeling time was negatively correlated with the number of incorrect feature terminations. This would be expected as explained above. However, it was not expected that a negative correlation was shown between the modeling time and the number of reference geometry features. There was a negative correlation between the number of features and the number of sketch and feature segments per feature. Because the more complex features are used, the less number of features would have been required. Because only a few students completed the alteration task, no t-test and correlations were analyzed. For future reference, the attributes and derived quantities for the altered models were still evaluated as shown in Tab. 5. The feature retention percentage of the models originally created by the alteration group was much higher than the models originally made by the speed group. This indicates that the quality of the original models made by the alteration group is better than that of the speed group. This makes the models easier to be altered later. Speed Alteration t Significance (1-tailed) Total Students Participating 8 8 N/A N/A Students Completing Exercise 4 4 N/A N/A 69.6 71.5 0.14 0.446 Sketch Plane 1 1 - - Origin 0 0 - - Base Feature 1 1 - - Group Original Design Time Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 16 Orientation 1 Correct Feature Sequence 1 - - 0 0 - - Number of Features 19.25 17.50 -0.70 0.256 Reference Geometry 0.50 0.75 0.45 0.335 Average Number of Segments 8.47 4.74 -0.84 0.216 Incorrect Feature Terminations 6.50 6.25 -0.19 0.426 Number of Mirrors 0 0 - - Number of Patterns 1.00 0.00 -1.41 1.943 Tab. 3: Statistical t-test between original model attributes. Time Orientation No. of Features Reference Geometry Avg. No. of Sgmts 0.127 0.280 -0.022 -0.484 -0.467 0.425 0.764 0.502 0.960 0.224 0.243 0.294 0.399 -0.747 0.058 0.330 0.189 0.328 0.033 0.891 0.425 0.654 -0.274 -0.227 0.468 0.371 0.512 0.588 0.242 0.365 -0.343 -0.328 0.269 0.405 0.428 0.519 -0.205 -0.430 Orientation No. of Features Reference Geometry Incorrect No. of Feat. Term. Patterns Avg. No. of Sgmts 0.626 0.287 -0.119 Incorrect Feat. Term. 0.779 Bold values are significant at the p<0.1 level. Tab. 4: Correlations between attributes of the original models. Speed Alteration Alteration Time 87 63.3 No. of Features Parent Model 24 25 Reference Geometry 1 1 Avg. No. of Sgmts. 4 4.06 Incorrect Feat. Term. 9 9 No. of Mirrors 0 0 No. of Patterns 0 0.5 No. of New Feat. 18 16.5 No. of Feature Deleted 10 11.5 Percentage Retention 37.5 51.65 Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 17 Retained Without Change 0 0 No. of Feature Changed 6 8.5 Tab. 5: Average results between altered model attributes. To compare the perception of what the students believe would be helpful during alteration with the above empirical results, survey data for the exercise tabulated by incentive groups are shown in Tab. 6. The attributes were rated on a seven point scale with a seven signifying that this attribute would be very helpful. There was a very small difference in opinions between the incentive groups. The students preferred the complex feature to the simpler features. Both groups preferred using patterns and mirrors in replicating geometry. The students had neutral opinions regarding the naming of features and the use of datum geometry. All three ratings (intuitiveness of organization, intuitiveness of feature order, and overall model quality) were higher for the models created for ease of alteration. These models were perceived as better structured, easier to understand, and better design intent conveyed. The survey results are generally in agreement with the analysis data. Speed Alteration t Significance (1-tailed) Naming Features 5.38 4.83 0.850 0.206 More Complex Features 4.50 4.33 0.197 0.423 Simpler Features 4.13 3.67 0.432 0.337 Patterns and Relations 5.25 5.83 -0.926 0.186 Mirror and Copy 5.50 5.67 -0.231 0.410 Referencing Datum Planes 4.88 4.83 0.042 0.483 Intuitive Organization 5.00 5.13 0.195 0.424 Intuitive Feature Order 4.17 5.00 0.995 0.170 Overall Rating 4.50 5.50 1.143 0.138 Parent Model Tab. 6: Survey results. 5 DISCUSSION The modeling and alteration exercise provided some valuable findings in understanding the CAD modeling procedures of novice CAD users. Generally the students did not know how to set the correct origin and coordinate system. The correct coordinate system is the very important first step to build a model which can capture the design intents. The students had little concept of choosing the correct feature terminations and feature sequence. For example, when designing a “through hole”, most of the students would set the depth of the hole rather than to set a “through surface”. Some feature operations, such as “edge blend” and “chamfer”, should be designed after all the main features are designed. Most of the students are not familiar with using “mirror” and “pattern” (named as “instance” in NX), although they agreed in the survey that these features are very helpful. The work also provided some empirical evidence on relating the model attributes to design intent. It was found that original modeling time was positively correlated with the number of features and negatively correlated with the number of complex feature. These are in agreement with the previous works [6], [9]. Using the reference features and defining correct feature terminations will cause more modeling time but will make the models easier to be altered. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 18 There are some limitations of this work. One significant limitation is that the sample size of the student participants was very small. Because of this, the correlations of the model attributes for the altered models were not analyzed in this work. Another limitation is that the participants were all freshman students. They had very limited experience on CAD. They might not know any alternative ways in which they could design and alter a model. 6 CONCLUSIONS AND FUTURE WORK This paper presented a modeling and alteration exercise that was implemented in a freshman class. The CAD program NX was used. The work was to assess the CAD modeling procedures of novice CAD users. Two groups of students were formed and incentivized with different goals. The first group’s goal was to design the part as quickly as possible. The second group’s goal was to design the part so that it can be easily altered. The parts were then exchanged between the groups and altered. A survey was conducted which assessed the students’ perception of the original models. The attributes of the completed original and altered CAD models were analyzed. The analysis data and survey results are presented in this paper. The preliminary results showed that the incentive has an effect on the completion time of CAD modeling. The completion time and rate shows that the models originally created by the alteration group were easier to alter than the models created by the speed group. The results also show how the feature selection and organization correlate to proper design intent. This work provided base data for the study to form preferred CAD modeling procedures. The preferred modeling procedures can be adopted by CAD educators to instruct students to create CAD models more effectively and efficiently. Future work will be focused on involving more students and more expert CAD users in the exercise. The engineers working in the industries will be engaged in the experiments. With more data collected, more correlations between the model attributions and design intent will be found. All the data collected on different CAD platforms (including Pro/Engineering, Solidworks, and NX) will be compared. They will be used to form a “generic approach” for conveying design intent in CAD modeling. This “generic approach” and the concept of design intent will then be taught to the students. ACKNOWLEDGEMENTS The work is supported by the National Science Foundation under EEC Grant Numbers 1129403 and 1129411. Any opinions, findings, conclusions, or recommendations presented are those of the authors and do not necessarily reflect the views of the National Science Foundation. The authors also thank the PACE program (Partners for the Advancement of Collaborative Engineering Education) for providing NX software. REFERENCES [1] [2] Alducin-Quintero, G.; Contero, M.; Martín-Gutiérrez, J.; Guerra-Zubiaga, D. A.; Johnson, M.: Productivity improvement by using social-annotations about design intent in CAD modelling process, Online Communities and Social Computing (Editors: Ozok, A. A.; Zaphiris, P.), Springer Berlin Heidelberg, 2011, 153-161. Bhavnani, S. K.; John, B. E.; Flemming, U.: The strategic use of CAD: an empirically inspired, theory-based course, in the Proceedings of the Human Factors in Computing Systems Conference, Pittsburgh, PA, 1999, 183–190. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com 19 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] Chester, I.: Teaching for CAD expertise, International Journal of Technology and Design Education, 17(1), 2007, 23–35. Dankwort, C. W.; Weidlich, R.: Guenther, B.; Blaurock, J. E.: Engineers’ CAx education – it’s not only CAD, Computer-Aided Design, 36(14), 2004, 1439-1450. David, R.; Frischknecht, A.; Jensen, C. G.; Blotter, J.; Maynes, D.: Contextual learning of CAx tools within a fundamental mechanical engineering curricula, in the Proceedings of PACE (Partners for the Advancement of Collaborative Engineering Education) Forum, Provo, UT, July, 2006. Hamade, R. F.; Artail, H. A.; Jaber, M. Y.: Evaluating the learning process of mechanical CAD students, Computers & Education, 49(3), 2007, 640–661. Iyer, G. R.; Mills, J. J: Design intent in 2D CAD: definition and survey, Computer-Aided Design and Applications, 3(1–4), 2006, 259-267. Johnson, M. D.; Diwakaran, R. P.: Examining the effects of CAD model attributes on alteration time and procedure, in the Proceedings of the ASME International Design Engineering Technical Conferences, Montreal, Canada, August 15-18, 2010. Johnson, M. D.; Diwakaran, R. P.: An educational exercise examining the role of model attributes on the creation and alteration of CAD models, Computers & Education, 57(2), 2011, 1749-1761. Kitto, K. L.: The role of CAE tools in engineering technology, in the Proceedings of Frontiers in Education Conference, 1993, 226-231. Lang, G. T.; Eberts, R. E.; Gabel, M. G.; Barash, M. M.: Extracting and using procedural knowledge in a CAD task, IEEE Transactions on Engineering Management, 38(3), 1991, 257–268. Ong, S. K.; Nee, A. Y. C.; Xu, Q. L.: Design Reuse in Product Development Modeling, Analysis, and Optimization, World Scientific Publishing Co., River Edge, NJ, 2008. Rynne, A.; Gaughran, W. F.; McNamara, B.: Parametric modelling training strategies to capture design intent, in the Proceedings of 17th International Conference on Production Research, Blacksburg, VA, 2003. Rynne, A.; Gaughran, W.: Cognitive modeling strategies for optimum design intent in parametric modeling, Computers in Education Journal, 18(1), 2008, 55–68. Toogood, R.: Pro|Engineer Wildfire 2.0 Tutorial and Multimedia CD, SDC Publications, 2004. Wu, J. C.: A Study of the learning models employed by industrial design students when learning to use 3D Computer-Aided Design (CAD) software, the International Journal of Arts Education, 7(1), 2009, 200-228. Ye, X.; Peng, W.; Chen, Z.; Cai, Y.: Today’s students, tomorrow’s engineers: an industrial perspective on CAD education, Computer-Aided Design, 36(14), 2004, 1451-1460. Computer-Aided Design & Applications, PACE (2), 2012, 9-19 © 2012 CAD Solutions, LLC, http://www.cadanda.com