25 Oct 2013 |
article de recherche |
Technologies de l'information et des communications
Maintaining Consistency between CAD Elements in Collaborative Design using Association Management and Propagation: Research Paper Introduction
[highlight]Un “Research Paper Introduction” est un billet présentant une introduction d’un article de recherche réalisé par des chercheurs de l’École de technologie supérieure (ÉTS) de Montréal, Québec, Canada. [/highlight]
The complete definition of a product often requires the collaboration of various partners. Data sharing and exchange between partners has thus become an important task throughout a product’s entire life cycle. Even while subsets of the product definition are exchanged (as work packages) and modified by various partners, the global product definition must remain consistent. The research paper focuses on maintaining consistency between Computer-Aided Design (CAD) work packages and the global product digital mock-up (DMU). The approach is designed to ensure better management of the associations between objects when a work package is extracted from the global DMU, modified by a partner, sent back to the originator and then re-inserted into the global DMU, which must be modified in turn so as to maintain consistency (fig. 1).
Work Package Evolution
Whatever the motivation behind the work package evolution, it is expressed through CAD object versions as modifications to geometry. The original object (O-Object) being sent from an original equipment manufacturer (OEM) to a partner as part of an iWP will differ from the next version of the object being returned (R-Object) to the OEM by the partner. Each R-Object is either an evolution of a previous version or it is a completely new version. The objective here is to associate, or to establish correspondence between entities that belong to the O- and R-Objects. We distinguish three cases. To simplify the discussion, objects are hereafter limited to faces.
Case 1 – Valid associations: In this case, the matching between entities of the R-Object and those of the original O-Object is obvious if all the entities’ ID’s are preserved (fig. 2).
Case 2 – Partially undetermined associations: In this case, some entities are orphans upon first analysis. This occurs either because the number of entities differs between the O- and the R-Objects, or because some entities find no ‘trivial’ equivalent via geometric equivalence or ID persistence. The former happens when m entities of the O-Object are destroyed and replaced by n entities on the R-Object. (fig. 3)
Case 3 – Undetermined associations: In this case, all entities’ IDs are destroyed and their cardinality may vary as well. This may occur when there is a passage through a neutral format or when the entire object is replaced by a new one. All constituting entities are therefore considered to be new. (fig. 4).
Figure 5 summarizes the three levels of complexity in reconciling associations and propagating changes, according to the work package change scenario. Recall that the objective is not to eliminate user intervention but rather to assist the user via a management model.
Association management model (DMU – AMM)
The association management model (DMU – AMM) enables digital mock-up/work packages to be reconciled after their modification and assist the user in propagating the modifications made. The association management model operating mechanism works in three steps:
- Capture of the initial associations between the initial work package (iWP) extracted from the DMU and then sent to a partner (or partners), and the initial state of the OEM digital mock-up (iDMU);
- Identification of the correspondence between elements of the initial work package (iWP) and those of the modified work package (mWP), as per three cases discussed above. This mapping enables the transposing of the associations captured between the iDMU and the iWP to the mWP, thereby expressing the reconciled associations;
- Reconciliation of the associations between the modified work package (mWP) and the initial digital mock-up is used to propagate changes to the DMU to obtain a modified DMU (mDMU) and maintain consistent information.
The inputs to this model are the initial digital mock-up (iDMU), the initial work package (iWP) and the modified work package (mWP). The correspondence algorithm is used to establish the correspondence between entities of the initial and modified work packages and to transpose the associations from entities of the iDMU towards the modified work package. The model output is the modified mock-up associated to the modified work-package (fig. 6).
From Initial associations to Change propagation
The proposed mathematical representation enables modeling of the relevant associations between iDMU, iWP and mWP, so as to propagate changes to the DMU (Fig. 7), with user involvement, in order to ensure geometric consistency. Prior to its extraction, the iWP is associated with the iDMU. The Association matrix ([Ma]) is used to represent these initial associations. Next, the evolution of the iWP towards the mWP is analysed, using a correspondence algorithm, and corresponding entities are associated through a Correspondence matrix ([Mc]). Combining these matrices enables transposing of the initial associations between the iDMU and the iWP into reconciled associations between the iDMU and the mWP. We will thus manipulate three matrices: the association matrix [Ma], the correspondence matrix [Mc], and the reconciliation matrix [Mr].
Illustration and validation
The collaborative design process requires exchanging data between OEMs and partners who contribute to the evolution of product definition. Data must be reintegrated into the OEM DMU after modification by the partner. The steps of the collaborative process using the proposed Association Management Model (AMM) are summarized as follows (Fig. 8):
- Extract the iWP and its association data sheet (ADS) from the iDMU;
- Send the iWP to the partner;
- The partner adds value to the WP and returns the mWP to the OEM DMU;
- The DMU-AMM is used to capture initial associations (iDMU/iWP), identify the correspondence (iWP/mWP) and reconcile associations (iDMU/mWP);
- Propagate changes to the DMU (by the user) by using the reconciliation matrix and the association (iWP/iDMU) data sheet (ADS).
To illustrate and validate the DMU-AMM, a DMU made of a three-jaw chuck is considered (Fig. 9). The chuck itself belongs to a work package. The modification scenario covers two needs:
- A change in the technological solution: T shaped to dovetail slide linkage;
- Add texture to the front face so as to improve gripping.
- In the first step, the initial association matrix [Ma], between the iWP and the iDMU is extracted (as an Excel spreadsheet), as well as the ADS. The associations are assembly constraints established within the DMU prior to extraction of the WP. This step consists of scanning the various assembly constraints to identify the entities in the iDMU and in the iWP that are associated to one another;
- In the second step, correspondences between entities of the iWP and of the mWP are established. The entities of the mWP are reconciled through the correspondence algorithm. This enables the correspondence matrix [Mc] to be generated (as an Excel spreadsheet). This step is essential for the transposition of initial associations towards the mWP;
- The third step consists of reconciling the associations between the mWP and the mDMU. The reconciliation matrix [Mr] is calculated by multiplying the correspondence matrix [Mc] by the initial association matrix [Ma]. The reconciliation matrix is provided to the user in an Excel spreadsheet so that the user can make the necessary changes to the DMU in order to obtain consistent associations (Fig. 9) and control the validity of the changes according to the association data sheet. An mDMU entity will probably be modified if it has an association with an entity of the mWP that has been modified. Therefore, we provide the user with decision-making support, enabling him/her to make the changes as required to obtain consistent associations and geometry. The user is alerted with a message indicating which entities are most likely to be modified. An annotation is placed on each of them (Fig. 9).
This work was performed within an industry-led research program called the Collaborative development for Product Lifecycle Management, which aims at developing a collaborative environment for better information management in the product development process while maintaining confidential data security.
To understand more about Maintaining Consistency between CAD Elements in Collaborative Design using Association Management and Propagation, we invite you to read the Research Paper available in the Journal « Computers in Industry » at the following link:
Louhichi B. and L. Rivest (2013). Maintaining Consistency between CAD Elements in Collaborative Design using Association Management and Propagation. Computers in Industry, Available online 29 September 2013. PDF