New automotive manifold for the latest generation two cylinder engine.
The complete automotive manifold is composed of two single parts injection moulded in a glass reinforced Nylon (PA66_GF35), these parts are subsequently joined by vibration welding.
The development of the automotive manifold during the standard pressure testing showed the welding quality, at a higher test pressure than was the requirement of the customer, pointed to a potential problem. Observations of failed manifolds showed a non homogeneous welding seam, an aspect that would have create difficulties for the reliability of the product during its possible life time. This is based on the years of experience of Magneti Marelli in the field of plastic air manifold, and their high standard of production quality.
The automotive manifold after a burst test
Testing of tensile
It was decided to first do a comparative test of the strength of different areas of the welded seam. This was done in accordance with the standard procedure of Magneti Marelli. 14 areas were selected from which 14 samples were obtained for the tensile testing.
Tensile test of a sample of the automotive manifold
On the basis of the results achieved and analysing the two initial manifolds, subsequently processed by vibration welding it was observed that the two printed parts appeared to have problems of matching due to part distortion. The failure to create a homogeneous welding and the presence of excessive tensions in some areas of the welded seam influenced the resistance to stress of the product. The part would perform under normal operating pressure, but tests indicated the expected weld strength was below the engineered values expected.
It was decided to analyse through the use of the Moldflow software simulation, the injection molding process to define the corrective operations to reduce the part warpage and analyze the potential adjustments for part changes.
Simulation of the deformations on the cover
The software calculation showed a possible warped geometry and demonstrated how in some areas the deviations involved directly affecting the welding seam. Before changing the molding process it is necessary to calibrate the Moldflow simulation model of the current production test (named Test 1) both in terms of parameters and in the display of the warp occurring in the parts, comparing the parts obtained from the injection molding processes first test settings.
The realization of 3D mathematical models of parts to be compared with the calculated models was possible due to the technology of Reverse Engineering and to the consultancy provided by CRP Technology. The hardware tools and software used in Reverse Engineering allows the operators to obtain the surfaces of the produced parts and compare them with the CAD original model and generating report of comparison able to underline the differences in dimensions due to the productive cycle.
In particular in the molding of plastic injection parts with shapes that are more difficult to detect and to quantify with the standard equipment (calibres, altimeters, ecc…) it is the dimensional shrinkage of the real part and the distortions that can be generated in the molded part during the cooling phase.
Two parts provided by Magneti Marelli were digitalized using an optical scanning system mounted on a 7 axes measuring machine. The combined use of the 7 axes machine and the laser system allow optical scanning simultaneous inside the same reference system.
3D optical scanning
The scanning procedure involves the progressive creation of point clouds that are lined up in the same reference system describing the object to being scanned.
The body and the cover: scanning completed
The complete scanning of the part is composed of more than 7,000,000 points.
An information so important is able to transfer the real object in a 3D virtual environment. A traditional coordinate measuring machine, is able to detect in short time a few hundred points on an object. These are often located on different sections which were previously chosen. These 3D point locations may lack enough data to allow all surfaces to be accurately represented. An example being the surface distortion may not be easily found utilizing a limited point system to describe the geometry. Only through a more complete scan utilizing the laser system can a true surface be generated.
Once the acquisition of the point cloud data has been completed, the scanning of the two objects surface data is optimized and then imported into the CAD environment in order to be compared with the original models. This allows the direct comparison and inspection of the reverse engineered production part with the original CAD Models.
In the first phase the data must be aligned in the coordinate system of the original CAD models. This alignment can be done via different methods. The following optional procedures can be used:
- Best fit: This method allows the comparison of the two files while leaving the coordinate systems un constrained on the file being checked. The software can ovelap the surfaces of the part file and those of the CAD model. This algorithm of alignment is useful to check the general correspondence of the CAD original part.
- Datum system: these data are defined in two files (data file and reference file) the same information of reference (plains, axes, centres) that are considered fundamental for the mounting or the functioning of the part. The component is aligned according to CAD model.
In this phase of analysis of the automotive manifold CAB257, a best-fit alignment has been preferred by Magneti Marelli.
From the analysis it was possible to see that there were distortions in the molded parts. In the model representations a perfect overlap shows as a mixture of colors of the two objects when compared. The the predominance of one color them indicates a gap or overlap that needs to be corrected.
The image below shows clearly that the scanned molded part as having in the central area of the ducts a distinct bend due to the warpage, seen in Orange.
The body: STL alignement (orange) on file CAD (blue)
Next image shows that the printed part “cover” is curved towards the outer part in all its surfaces near the entrance section of the plenum duct. Displayed in Yellow.
The cover: STL alignement (yellow) on CAD file (blue)
The differences can be displayed and quantified in short time mapping the surfaces of the parts detected with the color scale. Following image show the body that has dimensional difference within ± 2mm.
The body: mapping of the gaps detected
Next image show the “cover” that has dimensional differences within ± 1.5 mm.
The cover: mapping of the gaps detected
To verify the problem of a non-homogeneous welding between the body and the cover it is necessary to focus the attention on the welding line, the surfaces that are drown close to each other and melted through welding that have to create a weld bead. Next images underline the bigger deformations that can be generated on the weld line of the two printed components.
The body: deformation marker of the welding lines
The cover: marker mapping of the deformations on the welding lines
During the 3D testing phase the results have shown an explanation to the effects of the standard test of pressure resistance. The two printed parts are pushed one against the other on the welding machine. The deformations of the molded parts impede the correct functioning of the welding line contacting surfaces and for this reason, the distortions create a gap in contact or have only a small contact that causes an incomplete weld. Following image shows how a part of the body that was not welded. This exactly match the deformation measured on the welding line.
Correspondence between the deformations and the lack of welding
Managing all the process of quality control in virtual environment, it is possible to have measuring report that concentrates the analysis of the results only in the areas of interest with a meaningful time saving. The two images below show the shape of the project that must have the bead weld: as the two parts are drown close, the interference area melts up to the creation of the bead.
Creation of comparison section: theoretical welding lines
Theoretical section of welding line
Repeating the same positioning with the real geometry of the two parts, it is possible to see in the following images that due to distortions the material of interference addressed to proceed with the welding that in this section, is not possible.
Creation of comparison section: welding lines detected
Real section of welding lines
An interference of project equal to 1,5 mm that has led to a gap of 0,4 mm on molded parts.
Definition of new injection molding parameters
Once the 3D testing phase was given to CRP Technology, the values of the real deformed part of the molded plastic pieces were used by Magneti Marelli to control the results of the software and Moldflow simulation, too correctly calibrate the models reaction to adjustments in molding parameters.
The new simulation suggested improvements related the process parameters and some operations on the mould.
Production of new parts
New injection molded parts were produced according to the adjustments seen in the Moldflow simulation called Test 2.
New stroke pressure test
After producing some new welded automotive manifolds the pressure test was repeated, finding an increase of about 1 bar of pressure vs previous tests.
Results of burst testing
New tensile test
The test of resistance was carried out and the improvement in performance of the critical parts was clearly evident. In some areas the resistance value was doubled and even tripled.
Results of comparative tensile test (2° setting)
FEM structural test
At the end a FEM calculation was carried out to test if the new values measured are in accordance with the structural resistance of the automotive manifold and without the measured deformations of the first parts. This test corresponded closely, and was seen as a positive confirmation of the test data.
FEM calculation of the theoretical resistance of the welding line
The multidisciplinary activity was an experience that allowed Magneti Marelli to introduce a new quality control methodology for evaluation of the welding lines and of their control of distortion. In the past the sole parameter of reference was the values of the pressure testing (data provided by the customer) used to test the initial parts.
The new condition is based on a survey structured through:
- pressure test
- tensile test of welding bead
In case the part might have anomalies as it happened with the automotive manifold CAB 257, object of the current study:
- Analysis through the moldflow calculation
- Comparison between calculated deformations and real deformations (reverse engineering)
- Definition of parameter calculation of the process
- Intervention on process parameters
- Realization of modified parts (improvement)
- Pressure test
- Tensile test of welding bead
- FEM structural calculation and testing between the fem calculation and resistance test
Reverse Engineering was shown to be a rapid and effective tool for testing the automotive parts obtained by the process of plastic injection molding. The 3D optical scanning and the subsequent process of testing in CAD environement were able to provide Magneti Marelli with the information that could be used to improve the manufacturing process. These adjustments leading to a stronger weld, and a part that will be more durable through out its service life.
Once problems of welding were solved of the two halves of the automotive manifold CAB257, Magneti Marelli gave CRP Technology the test manifold together with the corresponding accesories (pipings, injectors, sensors). The aim of this second activity is the measuring of the part and then checking of probable mounting interferences inside the engine compartment; this check was carried out at virtual level “digital mockup” environment.
Laser scanning of the completed automotive manifold
Checking in digital mockup
Reverse Engineering completes the CAD/CAE/CAM technologies and the DESIGN LOOP, the process is represented by the development of the concept, designing, engineering and production.