Dear all,
I am looking for information regarding the methodological aspects of the modeling process with CAD softwares, particularly in the aerospace industry. At the moment I am focusing my searches on strategies for modeling machined parts. Sadly I am afraid most of this knowledge and experience remains inside companies, for reasons I can understand. It is indeed very hard to find practical insights: most papers I found are often excessively academical and lack the details of industrial practices.
The most common approach I have seen so far, especially with CATIA, is to use a “design by machining features” multi-body approach. The part is modeled the same way it will be machined: starting with a body representing the blank part then subtracting through Boolean operations bodies representing machining volumes (the volumes of materials the milling cutter tool will remove). The floor and walls fillets of each pocket (caused by the cutter tool geometry) are modeled on the machining bodies instead of being modeled directly on the final part. I found out some aerospace OEM use “standard” geometries for these machining bodies i.e. they model specific pockets like 2.5/3-axis pockets, 5-axis pockets, etc which are then “combined” onto the model.
One advantage I see from this method is enforcing the manufacturability of the parts being designed. It might even be possible to accelerate the design/modeling process (and I guess some companies have implemented this) by creating a catalog/library of User-Defined features e.g. 2.5, 3, 5-axis pockets with proper geometry (fillets) and why not with fillets radius being automatically calculated depending on the pocket geometry (cutter tool diameter selection).
One drawback however might be the fact the machined part is modeled with an approach using “negatives” (removing material) while it might be interesting for designers to model the part based on functional specifications with positive features such as web, flanges, stiffeners, etc.
My concerns are about the ease of creating robust parametric models, where design/modeling can be accelerated through Knowledge-Based Engineering tools for example. I would like also to be able using “design in context” techniques such as using assembly skeletons where the geometry of the part is driven by external references to a skeleton part in the context of the assembly.
Within this framework, I don’t have for now a clear picture of what are the geometry inputs given to a designer when he starts a model and how far the automation of the design of machined parts can go e.g. modeling generic part templates. Also I am wondering if it’s possible to implement a digital continuity (re-use) between the models created during different design phases (different levels of maturity) e.g. a part being detailed from preliminary to detailed design.
As a side note I would like to mention a concrete example for the modeling of machined wing ribs I have seen from one aerospace OEM using wireframe geometries (basically a skeleton of flanges and stiffeners) as input elements. The machining pockets are still being modeled independently but are attached to the wireframe. It gave me the idea it could be combined with a methodology I have found in an open-access standard from the KÄRCHER company. It can be found at Link (under: Documents for Design Methodology >> Handbook for Methology). They call it the “Minus-Minus methodology” (pages 17 to 19). It is used here for molded parts but could be adapted for machined parts. The idea behind it would be to keep the concept of substracting machining volumes but building the model with positive features.
Sorry in advance for imperfect English. I hope my questions are not too vague and some of you will have some experience they can share.
Kind Regards,
faerosp
I am looking for information regarding the methodological aspects of the modeling process with CAD softwares, particularly in the aerospace industry. At the moment I am focusing my searches on strategies for modeling machined parts. Sadly I am afraid most of this knowledge and experience remains inside companies, for reasons I can understand. It is indeed very hard to find practical insights: most papers I found are often excessively academical and lack the details of industrial practices.
The most common approach I have seen so far, especially with CATIA, is to use a “design by machining features” multi-body approach. The part is modeled the same way it will be machined: starting with a body representing the blank part then subtracting through Boolean operations bodies representing machining volumes (the volumes of materials the milling cutter tool will remove). The floor and walls fillets of each pocket (caused by the cutter tool geometry) are modeled on the machining bodies instead of being modeled directly on the final part. I found out some aerospace OEM use “standard” geometries for these machining bodies i.e. they model specific pockets like 2.5/3-axis pockets, 5-axis pockets, etc which are then “combined” onto the model.
One advantage I see from this method is enforcing the manufacturability of the parts being designed. It might even be possible to accelerate the design/modeling process (and I guess some companies have implemented this) by creating a catalog/library of User-Defined features e.g. 2.5, 3, 5-axis pockets with proper geometry (fillets) and why not with fillets radius being automatically calculated depending on the pocket geometry (cutter tool diameter selection).
One drawback however might be the fact the machined part is modeled with an approach using “negatives” (removing material) while it might be interesting for designers to model the part based on functional specifications with positive features such as web, flanges, stiffeners, etc.
My concerns are about the ease of creating robust parametric models, where design/modeling can be accelerated through Knowledge-Based Engineering tools for example. I would like also to be able using “design in context” techniques such as using assembly skeletons where the geometry of the part is driven by external references to a skeleton part in the context of the assembly.
Within this framework, I don’t have for now a clear picture of what are the geometry inputs given to a designer when he starts a model and how far the automation of the design of machined parts can go e.g. modeling generic part templates. Also I am wondering if it’s possible to implement a digital continuity (re-use) between the models created during different design phases (different levels of maturity) e.g. a part being detailed from preliminary to detailed design.
As a side note I would like to mention a concrete example for the modeling of machined wing ribs I have seen from one aerospace OEM using wireframe geometries (basically a skeleton of flanges and stiffeners) as input elements. The machining pockets are still being modeled independently but are attached to the wireframe. It gave me the idea it could be combined with a methodology I have found in an open-access standard from the KÄRCHER company. It can be found at Link (under: Documents for Design Methodology >> Handbook for Methology). They call it the “Minus-Minus methodology” (pages 17 to 19). It is used here for molded parts but could be adapted for machined parts. The idea behind it would be to keep the concept of substracting machining volumes but building the model with positive features.
Sorry in advance for imperfect English. I hope my questions are not too vague and some of you will have some experience they can share.
Kind Regards,
faerosp
