Minimising machining distortions with support from adaptive fixtures and adjustment of the NC-code

MATERIAL OCH PRODUKTION

KOMPONENTTILLVERKNING

Minimising machining distortions with support

from adaptive fixtures and adjustment of the

NC-code

Mats Werke RISE IVF, Daniel Semere KTH, Håkan Dahlquist System 3R, Peter Friberg

ETP, Anders Wretland GKN Aerospace

Minimising machining distortions with support

from adaptive fixtures and adjustment of the

NC-code

Mats Werke RISE IVF, Daniel Semere KTH, Håkan

Dahlquist System 3R, Peter Friberg ETP, Anders Wretland

GKN Aerospace

Abstract

The aim was to develop and test an adaptive fixture concept with a modified hydraulic chuck from System 3R. The conclusion is that the chuck may be a good concept for adjusting machining distortions but further development is recommended. Further, a literature survey concerning responsive fixtures for CNC-machining is presented. In addition, a virtual approach for compensation of machining distortions was tested. A framework of LS-Dyna algorithms for springback compensation after sheet metal forming was tested on a use case. The study indicates that the concept can be applied for NC-code optimisation on solid geometries with a complex manufacturing process chain composed of e.g. forging, heat treatment and proceeding machining processes. The deformations after machining were notably reduced in this numerical survey. Also virtual concepts for calculation of material removal strategies and clamping force analysis is suggested and demonstrated. The project was carried out with the support of physical experiments with an adaptive fixture and simulation with FEM based on the contour method.

Key words: Machining distortions, Adaptive Fixture, FEM, Machining strategies

RISE Research Institutes of Sweden AB RISE Report 2020: P102403

ISBN: 2020

Content

2 Adaptive Fixture demonstrator ... 6

2.1 Objective of the experiment ... 6

2.2 Experiment ... 7

2.3 Experimental Plan ... 7

2.4 Discussion and conclusion ... 13

3 NC-code adjustment using LS-Dyna ... 14

3.1 Objective of the virtual test ... 14

3.2 Virtual test ... 14

3.3 Discussion and conclusion ... 18

4 Material removal and clamping calculations ... 19

4.1 Objective of the virtual test ... 19

4.2 Material removal analysis ... 19

4.3 Clamping force analysis ... 21

4.4 Discussion and conclusion ... 22

References ... 22

Appendix 1... 23

Appendix 2 ... 38

Appendix 3 ... 42

Preface

This report presents results from the SME Aeronautics project "Minimising machining distortions with support from Adaptive Fixtures" (Vinnova Dnr

2019-02881).

The project was carried out in collaboration between System 3R, ETP Transmission

systems, KTH Industrial Production, RISE IVF and GKN Aerospace Engine Systems

AB. Mats Werke (RISE IVF) was project leader and Peter Ottosson (RISE IVF)

contributed with chapter 3 “NC-Code adjustment using LS-Dyna”. Daniel Semere and

Jagathishvar Jayakumar (KTH) contributed with chapter 2 “Adaptive Fixture

demonstrator” and Appendix 2. Peter Friberg (ETP) contributed with expertise in

experiment development and analysis of adaptive fixtures. Håkan Dahlquist (System 3R)

contributed with expertise, and the hydraulic chuck and clamping devices for the

experiment in chapter 2. Anders Wretland (GKN Aerospace Engine Systems AB)

contributed with test case for NC-Code adjustment in chapter 3 and expertise and in

machining distortions. Yuyin Jin, a previous trainee student at RISE contributed with the

calculations of clamping forces in chapter 4 and Fredrik Werke who holds a B.Sc. in

Mechanical engineering from Chalmers University of Technology contributed with a

literature survey concerning adaptive fixtures in Appendix 1. The authors wish to

acknowledge Vinnova for their financial contribution to this project. The project was

carried out during the period 2019-06-01 - - 2020-12-31.

Corresponding authors:

mats.werke@ri.se

danielts@kth.se

Summary

This report describes an experiment with an adaptive fixture for surface milling of a steel plate and the results concerning capabilities for reducing machining distortions. In addition, virtual principles for NC code optimization, analysis of machining strategies and calculation of clamping forces are described. Further, a literature study concerning adaptive fixtures is presented in Appendix 1. The results are summarized below:

Adaptive Fixture: For distortions close to the corners of the plate, i.e., near the pin

contact area, the adaptive support can provide a compensation effect with proper adjustment. Accurate compensations can be obtained with accurate repositioning of the pin. For distortions spread away from the pin contact and caused by cutting and clamping forces, the compensation by adjusting the pin alone does not correct it. The demonstrated adjustable hydraulic chuck may be a good concept for adaptive fixturing support during machining but further development is recommended.

Virtual compensation: This study indicates that the framework for spring back

compensation of sheet metal forming in LS-Dyna can be applied on solid geometries with a complex manufacturing process chain composed of e.g. forging, heat treatment and proceeding machining processes. The deformations after machining were notably reduced in this numerical survey.

Material removal and clamping: Calculations illustrate concepts for analysis of material

removal and clamping strategies. The calculations compare Top-Bottom and Bottom-Top material removal processes. The accuracy, compared to measurements vary and an increased mesh density is recommended for future tests.

1 Background

During machining, internal residual stresses are released which cause residual deformations and geometry errors on the final component. These are difficult to predict and in practice, therefore wide tolerances and several machining steps are applied to avoid reprocessing or, at worst, scrap. By taking into account the residual stresses during process planning of machining, a significant improvement in dimensional accuracy can be achieved. These can be determined using the Contour method [1] after which deviations from nominal geometry in subsequent machining can be calculated. The contour method means that the component is cut up with e.g. wire electro discharge machining (w-EDM), after which the deviation of the cut surface from an ideal flat plane is measured using, for example, geometric optical measurement technology (GOM) or coordinate measuring machine (CMM). The cut surfaces are then pushed back to their ideal shape virtually using FEM, which in practice means that the residual stress state is recreated in the model.

The contour method maps out of plane stresses based on the planar sections. The method can be extended to the analysis of 3-dimensional stresses by adding sectioning cuts, for example, 45 degrees [2] and 90 degrees [3] from the first cut in each section. In the Vinnova project "PRIZE" (Dnr 2016-03303), a strategy for calculating the 3-dimensional stress state has been developed using the contour method in combination with iterative redistribution of measured out of plane stresses in several sections according to [4]. In the Vinnova project "CUBE" (dnr 2018-03981), this concept has been further developed and a strategy for adjusting the NC code for compensating these geometric deviations has been developed [5].

Figure 1 Contour method [1]

In addition to adjusted NC code, a flexible restraint of the workpiece may also be required with respect to the resuspension effects that lead to the observed measurement deviations.

Chapter 2 describes experiments with an adaptive fixture configuration and Chapter 3 describes how to adjust machining distortions using FE-simulations. Chapter 4 describes how to analyze material removal strategies and calculate clamping forces using FEM.

- Appendix 1 describes a literature review of responsive fixtures for CNC-machining. - Appendix 2 describes the experimental setup for Adaptive Fixture demonstrator. - Appendix 3 describes measurement protocol from the chapter 2 experiment. - Appendix 4 describes a product sheet for the ETP-Unigrip.

2 Adaptive Fixture demonstrator

The experimental setup is based on a modified hydraulic chuck from System 3R (Appendix 2). Also concepts for adjustable fixtures from ETP were discussed (Appendix 4) but these were not included in the experimental setup.

GFMS System 3R International AB in Vällingby is an SME with 192 employees but is also part of the George Fischer group. The company develops fixtures for precision machining. ETP Transmission AB in Linköping is an SME with 52 employees that develops tool holders and fasteners for precision machining.

Experimental test on the applicability of the adaptive fixture concept was carried out at KTH production engineering lab.

2.1 Objective of the experiment

The experiment was intended to demonstrate the applicability of an adaptive work holding system for surface milling of a steel plate. The plate is fixed by four supports of which one is adjustable to compensate for reference datum errors if any. In these experiments, the errors on the reference plate are those;

1. Either induced from the machining force, clamping and spring back effects after machining or

2. Deliberately induced deformations by using different locator positions. The latter is to be used if the deformation from the first cause is found to be insignificant.

Figure 3. Adjustable Support Used

2.2 Experiment

Two sets of experiments were carried out. Each with steel plates with different thickness: 1. Plate 1: 450 x 450 x 12 mm and

2. Plate 2: 420 x 420 x 6 mm.

The plates have been clamped using the pin supported work holding of which one of the pin positions is adjustable, i.e., one of the four corners has been clamped using the adaptive (adjustable) fixture while the remaining three corners are clamped using fixed locators as shown in figure 4.

Figure 4. Experiment Setup

Adjusting the height of the adjustable support to a required value was carried out using a Mitutoyo Dial Test Indicator. The Dial Indicator was further used to indicate the surface profile of the plates used.

2.3 Experimental Plan

At first, the surface is rough milled with all locators set to ‘zero’ position. Then the following steps were attempted to demonstrate the compensation effect of the adjustable pin.

Step 1. Lock the pin position Step 2. Probe the surface

Extra support Adjustable support Steel plate Adjustable support Static locators

“As received” to create a reference surface for subsequent verification of the achieved depth of cut.

Step 3. Mill the surface

Step 4. Probe the machined surface and areas close to the support columns “As machined” but still clamped to verify the achieved depth of cut. Step 5. Loosen the pin and the clamps and stay for a while

“As released” (with materialized distortions after material removal) Step 6. Probe the surface again

Step 7. Compare the difference between the two measurements.

The result includes distortions including machining tolerances.

Step 8. Adjust the pin to counter the observed error and clamp. Step 9. Mill and probe the machined surface

Step 10. Check if the errors are corrected or eliminated.

Figure 5. Pictorial Representation of the experiment plan

The probing was carried out using the machine inbuilt Renishaw probe. The measurement was carried in such a way that it includes both the machined surface, and the surfaces close to the support corners. The cutting tool insert used is SECO XCKX13T308R – ME10, which were mounted to an 8-edged 80 mm diameter tool holder.

Trial 1 with thick plate: Plate 1 (450 x 450 x 12 mm) was first rough milled (1 mm depth of cut). After that, the steps explained in the experimental plan were followed with an additional cut (1 mm depth of cut). It was observed that the difference between the probing in Step 4 and Step 6 is negligible.

Figure 6. Trial 1 coordinate system and measurements. Blue Points indicate the unmachined surface, while the red points indicate the machined surface

Then with different configuration that was by inducing error by raising the adaptive fixture to 3.87 mm deviation. This was done in order to test the effectiveness of the adjustable chuck for compensations, the distortion was induced with deliberate error on the adjustable pin that later will be corrected. With the +3.87 mm deviation on the adjustable locator side the machined surface is deformed as shown in figure 7.

Figure 7. The blue points indicate negligible error surface, and the red points refers to induced error surface

In the compensation stage, the pin is readjusted to counter the induced error, i.e., lower the pin position from the rest by the same magnitude. The probing results are as shown figure 8.

Trial 2 with thinner plate, 6 mm: At first the thinner plate, 420 x 420 x 6, was rough milled with 0.5 mm depth. Following the steps described earlier, the following measurement data is obtained.

Figure 9. Trial 2 coordinate system and measurements. Blue points indicate the surface before machining, while the red points indicate the machine surface.

Figure 10. Blue Points indicate the clamped surface, while the red points indicate the unclamped surface.

2.4 Discussion and conclusion

The following observations are made based on the milling tests (Appendix 2): 1. For distortions close to the corners of the plate, i.e., near the pin contact area, the

adaptive support can provide a compensation effect with proper adjustment. Accurate compensations can be obtained with accurate repositioning of the pin.

2. For distortions spread away from the pin contact and caused by cutting and clamping forces, the compensation by adjusting the pin alone does not correct it.

The demonstrated adjustable hydraulic chuck from System 3R is a technically viable concept for adaptive fixturing support during machining. Further development is recommended.

3 NC-code adjustment using LS-Dyna

3.1 Objective of the virtual test

This chapter presents a virtual test that may be the foundation for NC code adjustment. The methodology is demonstrated using a test case: A forged and machined aerospace component made from a high strength nickel-base alloy, see Figure 11.

The work piece was hot forged in 1100 C with a screw press followed by solution annealing and then machining in several setups to final form. The geometrical shape has a tendency to drift out of tolerance and twist around the axial direction. The twist was max 0.51 mm and min -0.69 mm. The ambition was to set up a model for prediction and adjustments of the drift using the following steps:

a) Predict the internal stresses before machining using the Contour method. b) Simulate the material removal and predict the machining distortions using FEM c) If the machining distortions are out of tolerance, then adjust the level of the material

removal and simulate again according to 2.

d) Repeat c until the machining distortions are within tolerances. e) Adjust the original NC code accordingly.

3.2 Virtual test

One concept for virtual compensation of machining distortions, using LS-Dyna is described below with focus on b – d. An attempt to explore the possibilities to use technology developed for compensation of spring back in sheet metal forming to reduce the deformations occurring from release of residual stresses during machining was made. This was done as a pure numerical study. To perform this adjustment operation the final part geometry needs to be assembled within the geometry of the forged specimen. Furthermore, it requires that the residual stress distribution within the entire geometry is somehow determined. In this case, the latter is done by repeated application of the contour method in several section cuts along the geometry.

The idea behind geometric spring back compensation is schematically illustrated in Figure 12. Here three stages of sheet metal forming are shown in the top of the figure. Below is a schematic illustration of a section with the desired part shape in purple, the nominal tool in dashed green and the resulting spring back when the forming tools are Figure 11 Workpiece before and after machining and illustrated twist

removed, in grey. To the right the compensation principle is illustrated, where the offset from spring back is mirrored in the tool surface to create a new tooling geometry displayed as a solid green line. This new tooling geometry will be used in the next iteration of the process. The process is iterated until the accuracy is within desired tolerance.

Figure 12 Schematic principle of the iterative compensation loop for sheet metal forming as used in the software LS-Dyna.

The designed part to be machined from the larger forged specimen is illustrated in Figure 13. The intention here is to compensate the designed shape to account for the distortion that occurs from machining due to the residual stresses that originates from the forging and subsequent heat treatments. The compensation is done in a way that takes all the deformation as occurring in a step, which obviously is a major simplification from how it is done in reality by gradual removal of the excess material. The overall purpose here in this demonstration of the concept however is to determine whether this principle can be useful at all for handling machining distortions.

Figure 13. Forged specimen before machining in the top of the figure and the geometry of the machined part below

It is the interface between the component and the forged specimen that needs to be adopted to account for the deformations occurring from the residual stresses. A geometry with a solid mesh is shown in Figure 14. A section cut through this geometry reveals how the final part (displayed in blue) is located in the forged specimen (brown).

Figure 14 Solid finite element mesh of forged part on the uppermost part of the figure and a section cut through the part below with a close-up to the left.

In the analogy with rigid tools in sheet metal forming it is the interface that corresponds to the rigid surface that is to be adjusted by the spring back displacements of the final part as an effect of the removal of the outer part. Here then this removal takes place in one step, which is a simplification. This indicate a new geometry for the component, which deviates from the intended design. The result of spring back without numerical compensation is shown in Figure 15.

Figure 15. Resulting deformations from spring back without any compensation. NB. Legend displays deformations orthogonal to the image in mm.

By using the LS-Dyna algorithm for spring back compensation, with the original interface as the geometry representing the rigid tool, it is possible to move the position of the nodes in the interface a corresponding distance in the opposite direction from the spring back. Here it is crucial that the nodes in the interface is connected between both the outer and inner geometry. As we here are dealing with a solid geometry rather than a thin shell structure as in sheet metal forming, we can do the compensation separately on the top and the bottom surface of the specimen. Figure 16 illustrates the workflow.

Figure 16. Adjustment strategy

Figure 17 displays the result of this type of two step compensation process subsequent spring back simulation. In the first step only the top surface is compensated. The compensation causes change for this top surface but there is also a slight difference for the bottom surface according to the upper right in Figure 17. In the second step compensation of the bottom surface is performed, see lower half of Figure 17. In the same scale as previous images the reduction of spring back is difficult to appreciate. Figure 18 therefore illustrates the final geometry deviation after adjusting both top and bottom surfaces where the min and max values in this legend is reduced compared with Figure 17.

Figure 17. Resulting deformations from spring back after both first and second iteration of

compensation (both top and bottom). NB. Legend displays deformations orthogonal to the image in mm.

Figure 18. Resulting deformations from spring back after second iteration of compensation (both top and bottom). NB. Legend displays deformations orthogonal to the image in mm

3.3 Discussion and conclusion

This study indicates that the framework for spring back compensation of sheet metal forming in LS-Dyna can be applied on solid geometries with a complex manufacturing process chain composed of forging, heat treatment and proceeding machining processes. The deformations after machining were notably reduced in this numerical survey. Below, some of the results from the test case are discussed:

- The material removal was applied using instantaneous material removal, which is a simplified procedure that does not include e.g. stresses induced from the cutting tool. - Furthermore, there are limitations in terms of how to handle undercuts and material

nonlinearities etc.

- In addition, there are also interesting questions that can be further explored such as remapping of stresses to account for element distortions occurring in the

compensation process.

- Scale factors can be explored in the compensation to mimic some intermediate stage in the machining process.

- Ultimately, the interplay with fixturing and the component in combination with geometry compensation to reduce machining distortions would be of interest.

4 Material removal and clamping

calculations

4.1 Objective of the virtual test

This chapter demonstrates possibilities to analyse the effects on geometric distortions due to various material removal strategies and clamping. The demonstrations and models are based on the test case described in Section 4. The workpiece bulk residual stress before machining was calculated using the Contour method [4, 5]. This was followed by a simplified machining simulation approach where the cutting path was replaced by the CAD-surfaces of the final component and where the proceeding material removal calculations were carried out using a simplified instantaneous material removal approach. The calculations were performed using LS-Dyna and were compared with measurements according to Figure 19. The boundary conditions for clamping were excluded from the model in section 4.1 and included in section 4.2.

Figure 19. Measurement of deformations using CMM [4]

4.2 Material removal analysis

The effects on geometric distortions from two material removal strategies, Top-Bottom and Bottom-Top removal, are demonstrated. The results are analysed in nodes I, II, III and IV according to Figure 20. The out-of plane displacements (y-displacements) for the two strategies are shown in Figure 21 and in Figure 22 the results are compared with a simple one step removal strategy as well as measurements using CMM.

Figure 20. Top-Bottom machining approach

Figure 21. Y-displacements of 4 selected nodes (I, II, III, IV)

4.3 Clamping force analysis

In this example the model was clamped in 10 nodes and the out-of plane reaction forces due to clamping were calculated according to a Top-Bottom material removal strategy, see Figure 23. The out of plane reaction forces are shown in Figure 24 and the displacements are illustrated in Figure 25.

Figure 23 Top-Bottom removal with clamping

Figure 25 Y displacement (mm), Top -Bottom removal without clamping (left) and with clamping (right)

4.4 Discussion and conclusion

The results are summarised below:

- Top material removal may have a general higher influence on y-displacements

compared to bottom removal, see Figure 21. Further investigations are recommended. - The final geometric distortions are the same for the analysed nodes regardless of

Top-Bottom or Top-Bottom-Top removal, see Figure 21 and Figure 22.

- The accuracy, compared to measurements, vary from very high for node III to very low for node IV, see Figure 22. An increased mesh density may improve the results. - The reaction forces are very high for 3, 4, 7 – 10 after top removal, see Figure 24.

References

[1] Johnsson G 2008, Residual stress measurements using the contour method, PhD thesis, University of Manchester School of materials, Manchester UK.

[2] DeWald, Hill, 2006, Multi-axial contour method for mapping residual Stresses in

continuously processed bodies. Experimental Mechanics 46 473–490

[3] Prime, Newbom, Balog, 2003, Quenching and cold-work residual stresses in aluminum

hand forgings: Contour method measurement and FEM prediction” Materials Science Forum

426-432 pp. 435-440.

[4] Werke, Wretland, Ottosson, Holmberg, Michael Machens , Semere, 2018, Geometric

distortion analysis using a combination of the contour method and machining simulation, 51st

CIRP conference of Manufacturing Systems, Stockholm

[5] Werke, Hossain, Semere, Wretland, 2019, Machining distortion analysis of aerospace

components using the Contour method, Aerospace Technology Congress, October 2019,

Appendix 1

A review of responsive fixtures for

CNC-machining

Fredrik Werke (B.Sc. Mechanical Engineering)

Chalmers University of Technology

fredrik.werke@gmail.com

Tel. 0725443637

Date: 2020-08-30

PREFACE

This technical report is an appointment of RISE IVF AB and it has been prepared within

frames for the Vinnova funded project “Minimising machining distortions with support

from adaptive fixtures” (Drn: 2019-02881). The report reviews different adaptive

fixture solutions for machining in order to compensate for residual stress relief and

distortions. It has been put together by Fredrik Werke who holds a B.Sc. in Mechanical

engineering from Chalmers University of Technology.

TABLE OF CONTENTS

1. INTRODUCTION………...…………26

2. STATIC FIXTURING………...……. 27

2.1 Traditional static fixturing devices………28

2.2 Mould fixtures with negative shape……….. 29

3. RESPONSIVE FIXTURING………...29

3.1 The INTEFIX project………...……..30

3.1.1 INTEFIX fixture………...………...30

3.2 Aerospace engine blade fixture………..33

3.3 Intelligent flexible system (IFS)……….35

References………...………...36

INTRODUCTION

One drawback of rigid fixturing is for example the geometric distortion following when

the component is released due to the reaction forces on the locators. The research on

responsive fixtures if well established and there exist several proposals and commercial

choices to ensure high quality after fixture release. There are three major challenges for

fixtures, namely handling vibrations to ensure surface roughness, handling clamping to

ensure high precision positioning, as well as handling clamping to ensure springback

and distortion. The very last topic will be covered in this report and three responsive

fixtures will be highlighted.

2. STATIC FIXTURING

Fixturing is a balancing act between rigidly holding the workpiece and resisting

deformation due to overapplication of clamping pressure. It also needs to resist

vibrations generating during milling operations. In traditional machining methods, the

workpiece is fixed tightly to ensure workpiece stability and accuracy during the whole

machining process. The deformation caused by residual stress is accumulated gradually

and released intensively after finishing process, and the total deformation of the

workpiece is always big so that the machining specification can not be satisfied. When

the workpiece residual stress is released, it tends to deform to reach a new balance.

When the deformation is restricted by the fixture, it imposes forces to it, called reaction

forces [1].

A machining process with static fixturing is shown in figure 1 below where the blade is

disassembled from the fixture to release stress. Fixing requires the repair of the datum

of the blade after each release. The blade has to change the machining tool several times

to repair the datum. This procedure requires additional machining time, thereby

decreasing machining efficiency. Furthermore, re-clamping does not always lead to

acceptable results in terms of workpiece shapes which satisfy the tolerances defined by

the customers.

Figure 1 Traditional machining process workflow [2]

In traditional machining processes, the locators should be placed as far as possible from

each other and envelope the mass center of workpiece to maintain stability. If locators

are placed too far from each other, workpiece deformation cannot be released

sufficiently. While if locators are placed too close to each other, the workpiece will be

unstable. Therefore, a proper locating principle is imperative to satisfy the requirements

of adaptive machining [2].

2.1 Traditional static fixturing devices

T-track clamp: One common static fixturing device is the T-track clamp which is a

portable choice that slides along the grooves of the mills bed. T-track clamps can be

used in a wide range of applications and they provide an extremely strong clamping

force that is extremely reliable when machining. They are simple to construct and

customize to fit any individual application whether it is sheet metal or stainless steel

machining application [3]. Furthermore they are simple to manufacture in your own job

shop as figure 3 below depicts. The downside to T-track clamps is that they can be

difficult to write programs for as the placement of the clamps can vary.

Figure 2 T-track clamp [3] Figure 3 Custom made clamps

Vise: Another common fixturing method is to use a vise according to figure 4 below. A

vise is a set of two jaws with one jaw moving along a set screw. When a part is placed

between the jaws the set screw is tightened to hold it in place. Vises are a secure method

for holding parts and only apply clamping force along a singular axis. The downside to

using a vise is that parts fixtured in the vice usually need to have a flat bottom surface

and a predominantly flat side surface to properly hold the part for milling [3].

Figure 4 Machining Vise [3]

Traditional fixtures like t-tracks and vise both have in common that they are cheap and

robust.

2.2 Mould fixtures with negative shape

Mould fixtures with vacuum clamping are especially useful for thin walled parts.

Bending and scratching is avoided. The mould concept is hallmarked by the negative

shape relatively the part for full-surface clamping using vacuum and this is often

utilised in the industry according to figure 5 below. The main disadvantage of this

solution is high price and also the manufacturing process is practically non-flexible

since the moulds must be replaced or modified every time the workpiece changes [4].

Figure 5 Mould fixture with negative shape of the dedicated component [4]

Usually the vacuum clamping device is made from aluminum. The vacuum clamping

device is directly bolted to the table of the CNC-machine. Vacuum is applied between

the workpiece and clamping device. The atmospheric pressure forces the workpiece

down to the clamping device. This also secures no marks from clamping, as common

with mechanical clamping devices. Additionally no distortion of the workpiece is

created because atmospheric pressure is applied to the whole surface [4].

3. RESPONSIVE FIXTURING

There exist several synonyms for responsive fixtures like reconfigurable, intelligent,

adaptive, floating etc. However they all have in common the principle of closed loop

control with sensors and actuators. They automatically permit residual stress to be

continuously released and redistributed so component distortion is therefore reduced.

Responsive fixturing could be divided into adaptive clamping force control, adaptive

clamp location control, and also a combination of both. A conceptual illustration of the

machining workflow and control process using the adaptive fixture technology is

provided in figure 6 below. Unstressed clamping avoids the location error induced by

the deviation of the location datum [2].

Figure 6 Machining process workflow using responsive fixture [2]

3.1 The Intefix project

The INTEFIX project is an acronym of Intelligent fixtures for the machining of low

rigidity components. It was a three year EU funded project focusing on developing

flexible fixtures to minimize machining distortions. One project result from this

partnership considering case studies regarding machining distortion and clamping

pre-deformation will be presented further in this report and will be highlighted with the

INTEFIX symbol. Figure 7 below summarises duration, budget and project partners of

the project [1].

Figure 7 Intefix project partnership with duration and budget [1]

The main outcome of the INTEFIX project is the integration of new and state-of-the-art

technologies like sensors, actuators, control algorithms and simulation tools for

responsive fixturing [1].

3.1.1 INTEFIX fixture

This proposal concerns fixturing aeronautical structural parts manufactured in

aluminium alloy 7075-T7451 with a ribbed geometry by means of compensation of

residual stress. If a part should be manufactured out of a material for which the residual

stresses generated by the machining process itself would have a noteworthy effect on

the part distortion (e.g. Titanium alloys), the developed solution could not be applied

directly. The adaptive fixturing methodology follows a four step approach according to

table 1 below. In short the project has outlined one calculation software/engine and two

fixtures. The first fixture is for clamping the bottom side while the second one is for

sequentially clamping the upper side of the ribbed geometry.

Table 1 Sequence for the present methodology [1]

The first development is related to the methodology for the characterization of the

actual residual stress of the stock which is by the layer removal method according to

[1].

Figure 8 Residual stress in stock [1]

This residual stress profile is further used in the developed calculation engine that

predicts the final part distortion based on analytical formulas and thus not FEM.

Therefore the geometry is always simplified [1].

Figure 9 Residual stress profiles for different sections after the part bending [1]

After the calculation software has estimated the actual residual stress state of the stock

and the final part distortion is realized, a 2 step machining strategy (one on each side of

the part) that will yield the minimum part distortion is carried out. This strategy

indicates the required deformation δ to be applied on the centre of the part by the first

step fixture and the positioning of the part within the stock h, allowing for the

minimisation of the distortion of the part [1].

Figure 10 Parameters for the optimization of the machining

process [1]

The distortions are estimated by model calculations by entering the relevant parameters

according to GUI in figure 11 below and compensated by an active modification of the

stock for machining [1].

Figure 11 GUI for the control of the fixture and minimization of part distortion [1]

The first fixture according to figure 12 below bends the part according to the required

deformation δ to be applied on the centre of the part by the first step fixture and the

positioning of the part within the stock h, allowing for the minimisation of the

distortion of the part while optimising the machining time [1].

Figure 12 First step fixture [1]

The second step fixture according to figure 13 below would allow for the clamping of

the part after the ribbed geometry has been machined. The modules for holding the

ribbed geometry should clamp the different ribs in a simultaneous way for avoiding the

generation of uneven forces and to reduce the required clamping time. In order to avoid

the damage of the already machined geometry, the design of the clamping units for the

ribs is based on two self-balancing heads which assure that no lateral forces are induced

into the machined ribs [1].

Figure 13 Second step fixture [1]

These clamping units are hydraulically operated and mounted on modules for feeding

the different clamping heads with the hydraulic oil. The use of this hydraulic system

ensures that the final clamping forces are applied simultaneously on the whole part,

avoiding the generation of uneven forces on the part [1].

3.2 Aerospace engine blade fixture

Dongbo et al. proposes a flexible fixturing system for titanium aerospace engine blade

machining [5]. This fixture concept has been evaluated in several papers with respect to

performance and optimal clamping location [6] [7] [8] [9].

The blade body surface according to figure 14 below has a surface contour error of 0.08

mm after precision forging, which will lead to inaccurate positioning. The clamping part

is also only the blade body surface, whose stiffness is low due to the blade’s thin wall.

It´s maximum thickness is 5 mm, and the average thickness is 2 mm, so clamping

deformation is easily caused [5]. Therefore the two main process issues are positioning

and stiffness.

The fixture structure has three main parts, as shown in figure 15 below. 1) The fixture

base is used as a support base and is always connected to the machine tool; 2) the

positioning subassembly provides accurate positioning to the workpiece; and 3) the

clamping sub- assembly ensures stability. The clamping subassembly has two

independent clamping units, each with one actuator and two clamping heads to pressure

the blade body.

The main structural components of the fixture are made of steel, the clamping heads and

positioning heads are made of

PEEK-GF30 which is a relatively elastic

material [5].

Figure 15 Fixturing concept virtually and in CNC machine [5]

The clamping force is generated by the compression adjustment screw according to

figure 16 and figure 17 below. This force is transmitted to the active pressure plate via a

hinge structure, and finally transmitted to the active pressure head through two

independent hinges to the clamp block. This structure distributes pressure more

uniformly, and the pressures of the two active pressure heads are equal when ignoring

the frictional force [5].

Figure 16 Schematic clamping [5] Figure 17 Corresponding physical

structure [5]

This adaptive fixturing concept to overcome low stiffness involves a four-stage process.

Step one involves to calculate the overall deformation, maximum deformation, and

modal frequency of the blade-fixture structure by FEA.

Step two involves testing of the clamping forces and the corresponding deformations

via an eddy current sensor that measures the relative displacement variations.

Step three involves modal frequency analysis to determine the stiffness of the

blade-fixture system. This is an important indicator to evaluate a blade-fixture for the adaptive CNC

machining process.

In step four a cutting experiment and fixture application are used to verify the

feasibility of the fixture [5].

Figure 18 Stress on blade and on positioning heads Figure 19 Corresponding

deformation

To overcome positioning errors due to previous forging step, CMM measures of

component is conducted. The measurement model should be based on the measurement

points. The corresponding geometric parameters, such as chord length, and radii of

leading and trailing edges, should be calculated.

Those parameters are compared to the CAD model to obtain a geometric relationship

between them. Finally, the CNC machining model is reconstructed based on the

geometric relationship.

This fixturing concept and adaptive CNC machining technology has been proven

through engineering experiments to have a strong application value in processing

near-net-shaped thin-walled blades [5].

3.3 Intelligent flexible system

Zuperl et al. proposes a fixturing concept with on-line adjustment of clamping forces. It

adaptively adjusts the clamping forces as the position and the magnitude of the cutting

forces vary during machining. The closed loop clamping force adjustment is based on

reaction force stability criterion. When the workpiece residual stress is released, it tends

to deform to reach a new balance. When the deformation is restricted by the fixture, it

imposes forces to it, called reaction forces. Positive reaction forces at the locators

ensure that the workpiece maintains contact with all the locators from the beginning of

the cut to the end. A negative reaction force at the locator indicates that the workpiece is

no longer in contact with the corresponding locators and the fixturing system is

Figure 20 Control loop for responsive clamping force fixturing [10]

The sensing and clamping operations of the intelligent fixture are controlled by an

algorithm in a PC which is however not presented in the paper. The objective of the

force optimization algorithm is to minimize all the reaction forces. This could be

expressed as the minimization of the sum of the squares of the clamping and reaction

forces. Once the stability criterion is disturbed, clamping forces are adaptively adjusted

according to figure 21 below (upper). The fact that every reaction force is positive

according to figure 21 (lower) indicates that the workpiece will not detach from the

locators and the system is remaining stable [10].

Figure 21 Example of clamping forces and reaction forces

The average accuracy of the machined workpiece is improved by 12 % due to adaptive

control of the clamping forces. The present concept adjusts clamping location off-line

but future implementation of the system will incorporate also on-line repositioning of

the clamping elements [10].

References

[1] INTEFIX. Intelligent fixtures for the machining of low rigidity components - Project

results. https://www.springer.com/gp/book/9783319452906

[2] Jian-Hua, Yu. Zhi-Tong, C. Ze-Peng, J. (2015) A control process for machining

distortion by using an adaptive dual-sphere fixture. The International Journal of

Advanced Manufacturing Technology. Vol. 232(14) 2627–2640. DOI:

10.1177/0954405417699010

[3] Gregory, A. Lemire, D. Sengstaken, J. (2013) Reusable Conformable Workholding.

B.Sc. Thesis. Worcester Polytechnic Institute.

http://www.wpi.edu/academics/ugradstudies/project-learning.html

[4] Boehm feinmechanik (2020). Vacuum clamping devices

http://www.boehm-feinmechanik.com/html/vacuum_clamping_devices.html

[5] Dongbo, W. Hui, W. Peng, J. Zhang, K. Yu, J. Zheng, X. Chen, Y. (2019)

Machining fixture for adaptive CNC machining process of near-net-shaped jet engine

blade. Chinese Journal of Aeronautics and Astronautics. Vol. 257(9) 1947–26301. DOI:

10.1016/j.cja.2019.06.008

[6] Dongbo, W. Hui, W. Peng, J. Zhang, K. Yu, J. Zheng, X. Chen, Y. (2019) Analysis

of machining deformation for adaptive CNC machining technology of near-net-shaped

jet engine blade. The International Journal of Advanced Manufacturing Technology

Vol. 257(9) 1947–26301. DOI: 10.1007/s00170-019-03898-6

[7] Wang, H. Peng, J. Zhao, B. (2019) Modeling and performance analysis of

machining fixture for near-net-shaped jet engine blade. The international journal of

assembly technology and management. Vol. 257(9) 1947–26301. DOI:

10.1108/AA-08-2018-113

[8] Zhang, K. Dongbo, W. Wang, J. (2019) Research on Machining Fixture Layout

Optimization for Near-Net-Shaped Jet Engine Blade. Chinese Journal of Aeronautics

and Astronautics. Vol. 257(9) 1947–26301. DOI: 10.1016/j.cja.2019.06.008

[9] Dongbo, W. Zhang, K. Zheng, X. Chen, Y. (2019) Research on Machining Fixture

of Near-Net-Shaped Jet Engine Blade. Proceeding of the IOP Conf. Series: Materials

Science and Engineering

DOI: 10.1088/1757-899X/616/1/012004

[10] Zuperl , U. Cus, F. Balic, J. (2011). Real time control of clamping in an intelligent

fixturing system. Proceedings of the 22nd International DAAAM Symposium, Volume

Appendix 2

Description of adjustable hydraulic chuck

developed by System 3R

Figure 1

The experimental setup

Figure 2 Clamping support

Figure 4 Support parts manufactured by System 3R and Hydraulic chuck from System 3R

Figure 5 Modified Hydraulic chuck including spring for flexible movement of pin and position measurement device

Figure 6 Modified Hydraulic chuck with pin and position measurement device. Also Unigrip device from ETP is shown at the left. This entity was intended to be used for damping but was not included in the tests due to lack of time in this project.

Appendix 3

Results and data from Experiments with

Adaptive Fixture

Appendix 4

Through our international collaboration programmes with academia, industry, and the public sector, we ensure the competitiveness of the Swedish business community on an international level and contribute to a sustainable society. Our 2,800 employees support and promote all manner of innovative processes, and our roughly 100 testbeds and demonstration facilities are instrumental in developing the future-proofing of products, technologies, and services. RISE Research Institutes of Sweden is fully owned by the Swedish state.

I internationell samverkan med akademi, näringsliv och offentlig sektor bidrar vi till ett

konkurrenskraftigt näringsliv och ett hållbart samhälle. RISE 2 800 medarbetare driver och stöder alla typer av innovationsprocesser. Vi erbjuder ett 100-tal test- och demonstrationsmiljöer för framtidssäkra produkter, tekniker och tjänster. RISE Research Institutes of Sweden ägs av svenska staten.

RISE Research Institutes of Sweden AB Box 857, 501 15 Borås

Telephone: +46 10-516 50 00

E-mail: info@ri.se, Internet: www.ri.se

komponenttillverkning RISE Report 2020: P102403 ISBN: