SHRINKAGE, WARPAGE AND RESIDUAL STRESSES OF INJECTION MOLDED PARTS Tristan Koslowski, Christian Bonten, Institut für Kunststofftechnik, University of Stuttgart ,Germany Abstract The shrinkage and warpage behavior of injection molded parts is very complex. By using a new measuring method it was shown for the first time that the warpage can be classified as warping and distortion. In general, warpage can only be reduced substantially by a sufficiently high volumetric compensation in the holding pressure phase. Residual stress measurements show, deviating from the literature, that flow induced residual stresses can also induce compressive stresses in the part core due to pressure holding effects. Introduction Injection molding allows to produce geometrically complex parts at low cost in a single process step. [1] During and after production, changes in the produced part´s shape can occur, caused by shrinkage, volume contraction and thus warpage. [2] In many publications shrinkage and volume contraction unfortunately are not distinguished. The term “volume contraction” describes the volume decrease while the term “shrinkage” describes the polymer chains´ reorientation and thus relaxation while the volume keeps constant. This paper uses “contraction” and “shrinkage” in the above mentioned manner. The contraction and warpage behavior of the plastic parts has to be taken into account during the design and manufacturing phase of injection molds. Therefore, the mold shapes are generally made larger than the specified part dimensions. The prediction of contraction and warpage behavior is very complex. Nowadays contraction can already be predicted very well, but this is not true for warpage. The influence factors on warpage are still far less understood. Thus, very cost-intensive mold changes are often necessary until the specified part shape is achieved. State of Art The contraction and warpage of a plastic part are usually listed together in the literature. Contraction describes the dimensional deviation of a plastic part to the mold and is a normalized value, which can be determined for each plastic on standard specimen [3]. For warpage, however, the situation is different. In the German standard DIN 16742, warpage is defined as the sum of deviations in form, position and angle of the part due to a warping, distortion and contorting. An anisotropic contraction causes residual internal stresses in the part, which leads to warpage. [4] Due to the cooling of the part, a residual stress gradient as shown in Figure. 1 (top) occurs. After demolding, an anisotropic contraction over the wall thickness leads to a deformation of the part shape (Figure. 1, below). Thus a stress equilibrium is reached. [5, 6, 7] Figure 1. Residual stresses due to cooling effects. [5, 6] If constant contraction is present in all spatial directions, it can be assumed that only a very small warpage and residual stress occurs. This makes clear that the effects on the contraction behavior have to be considered separately before investigating the warpage. In this work, the different influencing factors on the contraction are distinguished into material-dependent, process-dependent and shape-dependent contraction. The material-dependent contraction behavior can be shown by means of a pvT-graph. The contraction of semi crystalline thermoplastics is higher than that of amorphous plastics because of the crystalline phase. The reason is a lower free volume in the solid state, which is caused by the tight crystalline arrangement of the polymer chains [8]. When the melt is under pressure, the free volume is reduced further. If additional fillers or reinforcing additives are added to the plastic, the pressure-dependence of pvT-behavior is less (Figure 2). thermal contraction of the individual layers (not mechanically coupled) averaged thermal contraction real thermal contraction (mechanically coupled) tensile stress compressive stress warpage SPE ANTEC ® Anaheim 2017 / 1620
Transcript
SHRINKAGE, WARPAGE AND RESIDUAL STRESSES OF INJECTION MOLDED PARTS
Tristan Koslowski, Christian Bonten, Institut für Kunststofftechnik, University of Stuttgart ,Germany
Abstract
The shrinkage and warpage behavior of injection molded parts is very complex. By using a new measuring method it was shown for the first time that the warpage can be classified as warping and distortion. In general, warpage can only be reduced substantially by a sufficiently high volumetric compensation in the holding pressure phase. Residual stress measurements show, deviating from the literature, that flow induced residual stresses can also induce compressive stresses in the part core due to pressure holding effects.
Introduction
Injection molding allows to produce geometrically complex parts at low cost in a single process step. [1] During and after production, changes in the produced part´s shape can occur, caused by shrinkage, volume contraction and thus warpage. [2] In many publications shrinkage and volume contraction unfortunately are not distinguished. The term “volume contraction” describes the volume decrease while the term “shrinkage” describes the polymer chains´ reorientation and thus relaxation while the volume keeps constant. This paper uses “contraction” and “shrinkage” in the above mentioned manner.
The contraction and warpage behavior of the plastic parts has to be taken into account during the design and manufacturing phase of injection molds. Therefore, the mold shapes are generally made larger than the specified part dimensions. The prediction of contraction and warpage behavior is very complex. Nowadays contraction can already be predicted very well, but this is not true for warpage.
The influence factors on warpage are still far less understood. Thus, very cost-intensive mold changes are often necessary until the specified part shape is achieved.
State of Art
The contraction and warpage of a plastic part are usually listed together in the literature.
Contraction describes the dimensional deviation of a plastic part to the mold and is a normalized value, which can be determined for each plastic on standard specimen [3]. For warpage, however, the situation is different. In the German standard DIN 16742, warpage is defined as the sum of deviations in form, position and angle of the part
due to a warping, distortion and contorting. An anisotropic contraction causes residual internal stresses in the part, which leads to warpage. [4]
Due to the cooling of the part, a residual stress gradient as shown in Figure. 1 (top) occurs. After demolding, an anisotropic contraction over the wall thickness leads to a deformation of the part shape (Figure. 1, below). Thus a stress equilibrium is reached. [5, 6, 7]
Figure 1. Residual stresses due to cooling effects. [5, 6]
If constant contraction is present in all spatial
directions, it can be assumed that only a very small warpage and residual stress occurs.
This makes clear that the effects on the contraction behavior have to be considered separately before investigating the warpage.
In this work, the different influencing factors on the contraction are distinguished into material-dependent, process-dependent and shape-dependent contraction.
The material-dependent contraction behavior can
be shown by means of a pvT-graph. The contraction of semi crystalline thermoplastics is higher than that of amorphous plastics because of the crystalline phase.
The reason is a lower free volume in the solid state, which is caused by the tight crystalline arrangement of the polymer chains [8].
When the melt is under pressure, the free volume is reduced further. If additional fillers or reinforcing additives are added to the plastic, the pressure-dependence of pvT-behavior is less (Figure 2).
thermal contraction of theindividual layers(not mechanically coupled)
averaged thermal contraction
real thermal contraction(mechanically coupled)
tensile stresscompressive stress warpage
SPE ANTEC® Anaheim 2017 / 1620
Figure 2. pvT-behavior of reinforced Polypropylene.
The process-dependent contraction is influenced by a large number of different factors. In addition to the injection mold design and the positions and shape of the filling points and cooling channels, the process parameters have to be considered. Each of these factors influence the material characteristics during the injection molding process. [9, 10, 11]
During injection, the plastic melt is exposed to high loads. The injection pressure, i. e. the pressure profile, is shown in Figure 3 (left). During the filling phase the pressure increases until the cavity is completely filled. Subsequently, during the holding pressure phase, the pressure is maintained to compensate the thermal volumetric contraction until the sealing gate point is reached. During the cooling phase, the cavity pressure decreases down to ambient pressure. The plastic temperature remains in the molten state during the filling phase and decreases with time down to the mold temperature (Fig. 3, left). [9, 10]
The pressure-dependent and temperature-dependent specific volume profile of an amorphous thermoplastic plastic during the injection molding process is shown in Figure 3 (right). Coming from the melt temperature, the specific volume decreases due to compression until the maximum cavity pressure is reached. During the holding-pressure phase, an isobaric curve progression occurs. When reaching the gate sealing point, the current pressure decreases down to ambient pressure, whereby the volume remains constant. Upon reaching ambient pressure, the plastic detaches from the mold wall and the volume contraction begins. The further cooling of the part takes place at constant ambient pressure down to room temperature. [8]
Figure 3. Process-dependent pressure and temperature.
The process-dependent influence on contraction is
especially pronounced, especially in the case of semi-
crystalline plastics. The volumetric contraction of injection-molded plates from the previously presented plastics (Figure. 2) is shown in Figure. 4. [10]
With an increasing content of fibers the volumetric contraction decreases. Likewise the unevenness, caused by a height deviation, as a characteristic value for the warpage of injection molded plates, decreases. [12]
This shows that unfilled semi-crystalline plastics in particular show the highest material-dependent contraction, which can be influenced by the process conditions to control the warpage and the residual stresses most efficiently.
Figure 4. Influence of fillers on contraction and warpage.
The shape of an injection-molded part has a decisive
influence on the contraction behavior of the plastic. Depending on the part shape and mold concept, different dimensions and tolerances can be reached.
In the withdrawn German standard DIN 16901 as well as in the current German standard DIN 16742, the shape-dependent contraction is divided into tool-specific dimensions and non-tool-specific dimensions [4]. Figure 5 (right) shows that this distribution is insufficient. The non-tool-specific dimensions must also be further differentiated into two types of contraction. Overall, there are three types of a shape dependent contraction. The free contraction only occurs in the thickness direction. In this case the plastic can contract unhindered. Longitudinally and transversally to the flow direction of the plastic there is an internal contraction hindrance due to already cooled edge layers. The result is a hindered contraction. An additional external contraction hindrance is caused by a contraction of the plastic on mold cores or undercuts. The freezing material is fixed in the mold and cannot contract further. This results in so called prevented contraction. [2]
For the investigation of the contraction and warpage behavior, a semi-crystalline polypropylene of the type BJ368MO from Borealis AG, Vienna Austria, was used. A low viscosity and good nucleation behavior ensures a homogeneous contraction and a symmetrical stress profile caused by cooling effects. Two plates of 60 mm x 60 mm were used: In Figure. 6 (left) the plate used for determining the contraction according to DIN EN ISO 294-3 is shown. An identical plate with additional ribs at the edges, which cause a prevented contraction, is shown on the right.
The experiment was carried out on an injection molding machine Arburg Allrounder of type 370S 700-100 / 70 from Arburg, Loßburg Germany, with a screw diameter Ø d = 25 mm.
The used mold was designed as an injection compression mold. This allows a simple variation of the part thickness t. For each thickness the height position of the diaphragm runner was adjusted accordingly (0.75 x t). In addition, two pT-sensors (p1, p2) were integrated for process monitoring.
Figure 6. Used part shape.
The experiments include all relevant process
parameters (Table 1). For this purpose a fractional factorial centrally composed design with 83 process variations was used.
The injection volume flow IV! of the machine (cm³/s) was chosen according to eq. 1. This allows to realize the same melt front speed fv in the cavity for all part thicknesses t. w is the width of the cavity. The variation of the demolding temperature eJ was achieved by means
of the residual cooling time ct according to eq. 2, where a is the temperature conductivity (mm²/s).
wtVv I
m ×=!
(1)
÷÷ø
öççè
æ
-
-××
×=
Moe
MoMec a
ttJJJJ
pp ˆ4ln2
2
(2)
Measurement Methods The contraction was determined according to the
standard. A measurement of the warpage according to the standard is not possible yet for the lack of a standard.
Till today, a standard measurement method for the warpage does not exist. Therefore, a new measurement method was developed with this work.
Table 1. Used process conditions.
Parameter Unit -α 0 α thickness t mm 1 2,5 5
melt temperature
ϑme °C 210 235 260
mold temperature
ϑmo °C 20 50 80
melt front speed
vf mm/s 150 500 850
holding pressure
ph bar 50 175 300
holding pressure time
th s 1 4,5 8
ejection temperature
ϑE °C 90 105 120
The height of the plate was determined tactilely at
nine measuring points (see Fig. 7). Thus, a characterization of the warpage as the maximum difference in height is possible.
In addition, the surface shape (eq. 3) or the surface profile in one direction (eq. 4) can be described with a polynomial function of second degree.
Figure 7. Warpage measurement method.
5432
212
0),( ayaxayaxyaxah yx +++++= (3)
cbxaxh x +×= 2
)( (4)
As a reference for the tactilely determined warpage
values, the warpage was additionally measured optically. For this purpose the plates were scanned with a 3D
scanner ATOS Core 135 from GOM, Braunschweig Germany.
Finally the internal stresses of the parts were measured in the center of the plate by the method of hole
60 m
m
60 mm
t 15 mm 2 m
m
t + 2
mm
p1p2
△hx,y
14
7
5
8
69
longitudinal
transversal
23
flow direction
SPE ANTEC® Anaheim 2017 / 1622
drilling according to standard ASTM E837-08 [13]. For this measurement method a small special designed cutter (diameter = 1.6 mm) drills gradually in into the plastic plate. The special design of the cutter allows a drilling without drill induced stresses. [14] Each drill step is 0.1 mm. After each step, the drill is removed from the plate and through the drilled holes, internal stresses are released which results in a change of the diameter of the hole. The in-plane strains are determined via strain gauges and the stresses are calculated in layers. For each plate shape, three measurements were implemented a few hours (<6 h) and again 24 h after production.
Results and Discussion
Contraction
The contraction of injection molded plastic parts
depends, as already mentioned, on a large number of influential factors.
The effect analysis of the process parameters on the volumetric contraction of the two plates is shown in Fig. 8. The graph reads as follows: When changing the value of a parameter by 1, the dependent variable changes by the value of the respective effect coefficient.
The total volumetric contraction of all process conditions consists of approx. 22–28 % longitudinal contraction, 18–24 % transversal contraction and 50–60 % thickness contraction. This means that the volumetric contraction is predominantly influenced by the thickness contraction.
It is found that the parameters of the holding pressure phase – which affect the compensation of the volume contraction – have a significant influence on the volumetric contraction (Figure 8). The influence of the process parameters is nearly identical for both plate shapes.
The hindered contraction is not fixated in the mold and therefore is influenced more strongly by the process parameters.
The effect analysis of the maximum height deviation
shows that it is influenced primarily by the parameters of the holding pressure phase, too. It can be assumed that the
warpage mainly results from an insufficient volumetric compensation. The plastic melt freezes on the exterior walls immediately and forms a dimensionally stable frame.
When the effect of holding pressure is insufficient (time and pressure), not enough plastic melt will flow into the mold cavity, resulting in the outer frame being deformed. Thus, increased warpage occurs.
Unlike the findings on contraction presented above, it can be seen that even though volumetric contraction is similar, the effects on the height deviation of the two shapes differ clearly. The prevented contraction shows much higher process dependence than the hindered contraction. The plastic is additionally fixed longitudinally and transversally to the flow direction.
This three-dimensional fixation of the plastic melt inside reacts more sensitively to the parameters of the holding pressure phase.
possible with the new measurement method. The simplified function (eq. 4) was used to evaluate the warpage. This function comprises only half as many factors as the surface function.
This new method allows a more precise classification of the warpage. a describes a warping (Figure 10, left), which is arched and can be mapped with a square function with high accuracy. On the other hand, b describes a distortion (picture 10, right) which is expressed by a twisted surface pattern. The final parameter c is a thickness value and depends completely on the thickness of the part.
Figure 10. Warping and distortion warpage.
The measuring points of the central process settings and the respective calculated surface profile according to eq. 4 as well the surface scan of the optical measuring
system are shown in Fig. 11. It can be seen that the simplified calculated surface profile describes the surface course very well. For the following investigations, an average of each of the three curves in longitudinal and transversal direction of the flow was used.
The new measuring method is explained in detail for the example of the holding pressure time at the central process settings of the experimental design. Figure 12, above, shows the maximum height deviation of the two plate shape at different holding pressure times. With an increasing holding pressure the height deviation decreases due to an improved volumetric compensation.
Figure 11. Comparison of the used measurement methods.
The parameters a and b (eq. 4) allow a more detailed view on the type of warpage (Fig. 12, center, below). In the case of hindered contraction the b-values longitudinal and transversal to the flow direction are almost zero (Fig. 12, below). Also the a-value transversal to the flow direction is very low. So the height deviation is caused by a parabolic warping ( a -Value) along the flow length. It is shown that an increasing holding pressure time reduces the value of a . Also the longitudinal contraction decreases due to the increased holding pressure time.
In the case of the prevented contraction Fig. 12, below shows very high b -values. Additional high a -values are shown in Fig. 12, center. This means that a warping and additional distorting appearances. The influence of the holding pressure time can be clearly demonstrated. Furthermore it can be seen that a is negative. That means that the deflection of the plate is not concave, like before, it is convex.
This shows that the warping of the hindered contraction occurs in opposite direction to the prevented contraction. This example shows the difficulty of the predictability of warpage. Despite a comparable contraction in both cases, the warpage behaves contrarily in its amount, shape and direction. Residual Stresses
The results of the incremental hole drilling method
show that the residual stresses of the hindered contraction along the flow direction have already developed the
idealized parabolic stress profile only a few hours after demolding (Fig. 13, left). There is only a slight shift in the stress profile (longitudinal) within the following hours (Fig. 13, right). Probably a pronounced stress relaxation took place along the flow direction. The contraction measurements confirm this assumption. The longitudinal contraction was higher than the transversal contraction.
This is particularly true in the case of semi-crystalline thermoplastics, where process-dependent polymer chain
Figure 12. Influence of the holding pressure time on the warpage. orientations along the flow direction have enough time to relax. The orientated chains fold together and cause a larger relaxation along the flow direction. [15] Transversal to the flow direction, lower stresses are present a few hours after production. An internal hindered contraction has been caused by edge layers which have already frozen. Thus, an increase of energy-elastic stresses occurs.
These energy-elastic stresses decrease with time. Presumably, the majority of the stresses are reduced due to a deformation of the part directly after demolding.
It is to be assumed that the high values for the in flow direction are caused by the higher stresses in flow direction at the point of demolding. Once the plastic part has assumed room temperature, the deformation will change only slightly. 24 h after production, the initial stress relaxation transversal to the flow direction is already completed to a large extent. The stresses longitudinally and transversally to the flow direction are almost identical.
The stress profile transversely to the flow direction in Figure 13, left, indicates flow induced residual stresses in the part core. [16] The volumetric compensation of the plastic melt by the applied holding pressure causes additional orientations in the core layer. These can completely relax within 24 hours so that the known parabolic cooling stress profile is present.
Figure 13. Residual stress caused by hindered contraction. The stress profile of the prevented contraction leads
to a significantly different shape after demolding (Figure 14, left). Although compressive stresses have resulted from cooling effects in the edge regions, these are substantially less.
The stress level in the part core in transversal direction as well as in longitudinal direction is significantly smaller and partly loaded under compressive stress. Compressive stresses in the mold core have been attributed to expansion stresses so far [17, 18]. In this case expansion stresses can be excluded. There was neither a mold deformation nor a demolding under residual pressure. It can be assumed that the low viscosity of the plastic allows a pressure transmission with hardly any pressure loss. Figure 15 shows the cavity pressure profiles measured by the sensor away from the gate as well as the ones by sensor near the gate.
Figure 14. Residual stress caused by prevented contraction.
The three-dimensional fixation of the plastic melt by the prevented contraction prevents a uniform relaxation of the polymer chains in the center of the part. It can be assumed that the flow induced residual stresses increase significantly due to the holding pressure effects. Presumably, the clearly process depending warpage results from this area. The small difference between the longitudinal and transversal stresses <6 h after production and the high distribution of the measured values will presumably cause the warpage by warping and distortion.
After a storage time of 24 h the stresses in the core relax, too. A parabolic stress profile is the result. The residual stresses in the edge layers are almost identical to the hindered contraction. The stress level in the part core is significantly lower.
Figure 15. Cavity pressure while part production.
Conclusions Within this work it was shown that the holding
pressure phase has a significant influence on the contraction and warpage behavior, especially of unfilled semi-crystalline plastics. Generally, a slight warpage results from a slight contraction. However the type of injection mold has to be taken into account.
Depending on the mold concept the contraction can occur as a free contraction, hindered contraction or prevented contraction. In addition to the dimensional deviations, the shape and orientation of the molded part can be altered by the warpage.
By using a new measuring method it was shown for the first time that the warpage can be classified as
warping and distortion. The prevented contraction, in particular, results in an increased warpage which depends on the process parameters markedly. In general, warpage can only be reduced substantially by a sufficiently high volumetric compensation in the holding pressure phase.
Although the same material and process conditions were prevailed, different residual stresses were measured in the injection molded parts. Residual stresses in mainly one direction cause warping. When residual stresses in two directions have an identical range, additional distortion occurs.
In the core of the part with prevented contraction, compressive stresses were present only a few hours after production. This shows that, deviating from the literature, flow induced residual stresses can also cause compressive stresses in the part core due to pressure holding effects. Demonstrably, these compressive stresses have relaxed within 24 hours.
References
1. C. Bonten Kunststofftechnik. Einführung und
Grundlagen, Hanser (2016). 2. W.B. Hoven-Nievelstein, Die Verarbeitungsschwin-
3. ASTM D955-08, Standard Test Method of Measuring Shrinkage from Mold Dimensions of Thermoplastics, ASTM International, West Conshohocken PA, (2014)
4. Deutsches Institut für Normung e.V., DIN 16742 Kunststoff-Formteile - Toleranzen und Abnahme-bedingungen, Beuth (2013).
5. S. Stitz, Analyse der Formteilbildung beim Spritzgießen von Plastomeren als Grundlage für die Prozesssteuerung, Dissertation, RWTH-Aachen, (1973).
6. P. Larpsuriyakul, H.-G. Fritz, Journal of Polymer Engineering and Science, 51, 3 (2011).
7. T. A. Osswald, International plastics handbook, Hanser (2006).
8. G. Menges, P. Thienel, Journal of Polymer Engineering and Science, 17 ,10 (1977).
9. D. Kusić, T. Kek, J.M. Slabe; et. al., Polymer Testing, 32, 3 (2013).
10. S. H. Tang, Y. J. Tan, S. M. Sapuan, et. al., Journal of Materials Processing Technology, 182, 1-3 (2007).
11. T. Lucyshyn, G. Knapp, M. Kipperer, et. al., Journal of Applied Polymer Science, 123, 2 (2012).
12 J. D. Santos, J. I. Fajardo, A. R. Cuji, et. al., Frontiers of Mechanical Engineering, 10, 3 (2015).
13 ASTM E837-08, Standard Test Method for Determining Residual Stresses by the Hole-Drilling Strain-Gage Method, ASTM International, West Conshohocken PA, (2009)
14 A. Nau, B. Scholtes, J. Nobre. Journal of Plastics Technology, 7 (2011).
15. B. Heise, H.-G. Kilian, G. Lüpke, et. al., Kolloid-Zeitschrift und Zeitschrift für Polymere, 250 (1972).
16. A. Guevara-Morales, U. Figueroa-López, Journal of Materials Science, 49, 3 (2014).
17. H. Domininghaus, P. Elsner, P. Eyrer, T. Hirth, Die Kunststoffe und ihre Eigenschaften, Springer (2005).
18. G. Wübken, Einfluss der Verarbeitungsbedingungen auf die innere Struktur thermoplastischer Spritzgussteile unter besonderer Berücksichtigung der Abkühlverhältnisse, Dissertation, RWTH-Aachen (1974).
