Magnetic coating for primary vehicle door seals with improved sealing performance.

Magnetic sealing has been widely used in the manufacture of refrigerator gasket seals for many years. Magnet strips are imbedded in an elastic extrusion, mostly flexible PVC. Numerous patents have been assigned to the design and assembly of magnetic refrigerator seals. For example, Okamura (ref. 1) claimed a patent for "Sealing apparatus for refrigerators and method of manufacture thereof," Merla (ref. 2) for "Sealing gasket between a door and its related shoulder on a cabinet, in particular of a refrigerator," and Gerdes et al (ref. 3) for "Refrigerator and method of gasket assembly construction."

Magnets have also been proposed in the fabrication of window seals for automotive vehicles. Mesnel (ref. 4) discloses the invention of an automotive window seal concept using an elastomer profile with an imbedded magnetic strip. White (ref. 5) discloses an apparatus for sealing a door, including a metal strip to be applied to the door and a sealing strip having a magnetic portion for application to the frame. Keys et al (ref. 6) claimed the invention of a magnetic window assembly for frameless and full flush window systems of automotive vehicles. The system consists of an elongated flexible-sealing strip that is mounted to the vehicle body and has a compressible body portion with a magnetic element embedded within. A complementary magnetic element is installed on the lateral edge of the window being sealed. Cittadini et al (ref. 7) claimed the invention of a magnetic gasket particularly suitable for forming a seal between a fixed part and a moveable part. The gasket consists of a profile and tubular seat housing a magnetic strip.

The concepts for magnetic sealing in the patents mentioned above all contain imbedded magnets in profiles. Their fabrication all have to involve in three-step processes. This not only increases processing complexity, but limits profile design versatility.

The ordinary primary door seal system for vehicles consists of a co-extruded dense carrier and sponge bulb tubular seal. The sponge bulb provides sealing functionality when compressed against the body frame. This type of seal is considered conventional, has been used for years, and generally performs satisfactorily. However, under some circumstances, a conventional design may not provide an adequate seal. For example, at high driving speeds, external air pressure may exceed the maximum seal force provided by sponge bulb and cause a seal failure.

To offset this problem, our proposal involves a new design of sealing profile created by incorporating magnetic rubber into a sponge bulb. This can be achieved by extruding either a thin layer of magnetic coating onto the sponge bulb, or a section of magnetic rubber into the sponge bulb. The magnetic coating or insertion will directly contact the body metal frame and therefore enhance the bulb sealing function.

This article deals mainly with the formulation and characterization of magnetic rubber to be used in the magnetic sealing system mentioned above. The magnetic rubber consists of an ethylene propylene diene rubber (EPDM) filled with a magnetizable ferrite powder (strontium type) at different loading levels. The effect of magnetizable powder, coupling agent and compatibilizer on compound curing characteristics, compound rheology and on magnetic force has been investigated.

Experimental

EPDM polymer was a high molecular grade with 8 wt% of ethylidene norbornene (ENB). Magnetizable powder was a strontium type with the formula of Sr[Fe.sub.12][O.sub.19]. The coupling agent used was neopenty(diallyl)oxy, tri(N-ethylenediamion) ethyl titanate. Compatibilizer was a maleic anhydride grafted EPDM (MA-EPDM).

Mixing of magnetizable powder with EPDM was performed in a lab-scale intermeshing batch mixer. Upside down mixing methods were used, and dump temperature was set at 330 [degrees] F. In the case of coupling agent and compatibilizer, they were added first with the EPDM and mixed for one minute followed by the rest of the ingredients. The total mixing time was set at 3.2 minutes. Curatives were added to the mixed batch on a 600 mm mill when the compound temperature was at 65 [degrees] C. Compounds were extruded into 25 mm wide by 2 mm thick ribbons using a rubber extruder set at 80 [degrees] C. Magnetization of the ribbon was performed with a proprietary magnetizer using a 2-pole configuration.

Compound viscosity and curing characteristics were characterized using a Mooney viscometer and a moving die rheometer according to ASTM D6204. Compound rheological properties were characterized using a Brabender PL2100 system in combination with a slot die. Test temperature was 80 [degrees] C. Apparent shear rate, apparent shear viscosity and shear stress were calculated by the Brabender software. Surface activity of magnetizable powder was measured by a FT-IR equipped with a SpliperiP micro ATR. PH value for the powder was measured according to ASTM D1512. Compound physical properties were measured according to ASTM D-412, D2240 and D624.

Results and discussion

Effect of magnetizable powder on compound curing characteristics

Table 1 lists the basic formulation used in this study. Magnetizable powder loading was varied from 0 to 500 parts per hundred parts of rubber (phr), while polymer loading was constant. Calcium carbonate was used to adjust overall filler content to keep polymer portion constant.

Table 1 - basic formulation

Ingredient     Parts per hundred   Ingredient       Phr
                of rubber (phr)

EPDM                  100          CaO               5
FEF black             100          Processing aid    1
Whiting            100 to 500      Paraffinic oil   120
Magnetizable                       Sulfur-70        1.6
  powder           100 to 500      TMTD-67           1
Zinc oxide            4.5          MBTS-67          1.2
Stearic acid           1           ZDMC-67          0.5
PEG                    2

Variation of Mooney viscosity and scorch time T5, with magnetizable powder loading is shown in figure 1. It is seen that Mooney viscosity decreases with magnetizable loading. This is due to the fact that magnetizable powder absorbs less oil than calcium carbonate. The higher the magnetizable powder loading, the less calcium carbonate in the formulation and the less oil it absorbs, resulting in lower compound viscosity. Magnetizable powder also acts to extend slightly the scorch time. This was believed to be partially caused by the acidic nature of magnetizable powder to retard curing. This was then confirmed when the pH value of this powder was tested and found to be between 5.3 and 6.4.

[GRAPH OMITTED]

Figure 2 shows the effect of powder loading on Ts2 and T90. It can be seen that the powder increases both values, which means it retards compound curing. This can also be understood by the slightly acidic nature of the powder. However, compared to the case of T5, the effect on Ts2 and T90 is more pronounced. An FTIR-ATR analysis of the powder surface revealed that the powder surface is very inert, not even a trace of water was detected on the powder surface, as shown in figure 3.

[GRAPHS OMITTED]

If water is present, it will show peaks in the wavenumbers of 3,300 [cm.sup-1] and 1,650 [cm.sup.-1], as shown in figure 4, the FT-IR scanning for water.

[GRAPH OMITTED]

This observation may indicate that Sr[Fe.sub.12][O.sub.19] (strontium ferrite) competes with ZnO in reacting with the MBT, a derivative from the decomposition of MBTS, to form a MBT-[Sr.sup.+] complex, as illustrated in the scheme shown in figure 5.

[ILLUSTRATION OMITTED]

It is thought that the [Sr.sup.++] reacts more quickly with MBT and forms more stable MBT-[Sr.sup.+] complex than [Zn.sup.++] due to its less stearic hindrance moiety of [Fe.sub.12][O.sub.19] compared to the stearic moiety. The stable MBT-[Sr.sup.+] retards further participation of the complex in crosslinking EPDM. More work is needed in this aspect to confirm the hypothesis.

Effect of coupling agent on compound curing and physical properties

A measurement of magnetic force for compounds shown in table 1 indicates that force magnitude is too weak to be used for any practical sealing profile design. It is therefore needed to increase the magnetic force by increasing magnetizable powder loading. Also, it was found that physical properties of magnetic compounds have been reduced with increasing magnetizable powder loading. To overcome this problem, coupling agent and compatibilizer were used at the dosages of 0.5% based on fillers and of 0.2% on polymer, as well as of 1% on fillers and of 0.2% on polymer. In the case of MA-EPDM, EPDM polymer in the formulation was reduced to take account of the EPDM portion of MA-EPDM.

It is seen, compared with the control, that the titanate coupling agent, at both dosages, lowers compound viscosity, retards scorching and slows curing. It also slightly reduces tensile strength and compound hardness and gives higher compression set values. However, it dramatically increases compound elongation. In the case of 0.5% titanate, it increases tensile strength and elongation after heat aging. In the case of 1% titanate, the trend is similar to that of the control, i.e, tensile strength and elongation are reduced after aging. This may be understood by considering the fact that [Ti.sup.+4] in the coupling agent acts as a free radical scavenger as [Ti.sup.+4] is reduced to [Ti.sup.+3] when it reacts with a radical (R*), as shown:

[Ti.sup.+4] + R* [right arrow] [Ti.sup.+3]R

It is also known that rubber vulcanization in the presence of TMTD/ZnO is a radical crosslinking mechanism (ref. 8). The reduction of free radicals by titanate will certainly reduce free radicals available for vulcanization and therefore reduce the curing speed. In the case of 0.5% titanate, the amount is probably not high enough to significantly reduce the free radical concentration. However, in the case of 1.0% titanate, the compound may have been overdosed and a significant amount of free radicals consumed, and not enough radicals would be available for vulcanization, resulting in an undercure situation. The coupling agent does act to reduce the viscosity and helps the compound retain physical properties after aging, without increasing the modulus or tensile strength. The reason for viscosity reduction by titanate may be that it interacts with carbon black in the formulation to improve carbon black dispersion and subsequently reduces compound viscosity as:

HO-carbon black-OH + RO-Ti-OR" [right arrow] R"O-Ti-O-carbon black-O-Ti-OR"

in the case of the MA-EPDM compatibilized formulation. Compared to the control, the compatibilizer gives not only higher compound Mooney viscosity, shorter scorch time, faster curing speed, but also higher modulus. This indicates that there is a strong interaction between MA-EPDM and the magnetizable powder. However, this strong interaction probably over reinforces the compound, reducing dramatically compound elongation.

Table 2 - effect of coupling agent and compatibilizer on compound
physical properties

Properties                                  Test    Control     0.5%
Compound                                   method             Titanate
characteristics

Mooney @ 121 [degrees] C, unit             D6204     51.6       42.6
T5 @ 121 [degrees] C, min.                 D6204     14.38       35
Ts2 @ 177 [degrees] C, min.                D6204     1.47       1.62
T90 @ 177 [degrees] C, min.                D6204     3.30       3.39

Physical properties

Tensile strength, MPa                      D412       3.0       2.6
Elongation, %                              D412       355       412
Hardness                                   D2240      77         73
100% Modulus, MPa                          D412       2.3       2.1
200% Modulus, MPa                          D412       2.2       1.9
300% Modulus, MPa                          D412       2.7       2.1
Tear strength, kN/m                        D624      18.5       19.2

Heat aged @ 70 [degrees] C/70hrs.

Tensile change, %                          D412      -3.3       +7.7
Elongation change, %                       D412      -7.3       +0.7
Hardness change, %                         D2204      +2         +2
100% modulus change, %                     D412      +4.3       +9.5
200% modulus change, %                     D412      +4.5      +10.5
300% modulus change, %                     D412      +3.7       +4.8
Compression set @ 70 [degrees] C/22hrs.,   D395B     26.9       26.8
  % (plied)

Properties                                    1%       0.5%       1%
Compound                                   Titanate   MA-EPDM   MA-EPDM
characteristics

Mooney @ 121 [degrees] C, unit               40.7      66.4      73.4
T5 @ 121 [degrees] C, min.                    44       13.26     14.17
Ts2 @ 177 [degrees] C, min.                  1.76      0.93      0.95
T90 @ 177 [degrees] C, min.                  3.42      2.76      2.86

Physical properties

Tensile strength, MPa                        2.9        4.1       5.0
Elongation, %                                526        259       197
Hardness                                      75        79        78
100% Modulus, MPa                            1.8        4.1       5.0
200% Modulus, MPa                            1.8        3.8       --
300% Modulus, MPa                            1.8        --        --
Tear strength, kN/m                          19.5      25.7      25.4

Heat aged @ 70 [degrees] C/70hrs.

Tensile change, %                            -3.4      +9.8      +4.0
Elongation change, %                        -12.4      -35.9     -22.8
Hardness change, %                            +2        +1        +2
100% modulus change, %                      +16.7      +4.9      +4.0
200% modulus change, %                      +11.1       --        --
300% modulus change, %                      +16.6       --        --
Compression set @ 70 [degrees] C/22hrs.,      36       21.8      20.2
  % (plied)

Effect of coupling agent and compatibilizer on compound rheology

A slit die was used to measure compound flow properties. The reason for such a choice was that shear rate generated in this die corresponds fairly closely to the shear rate experienced by this type of compound in actual extrusion. Figure 6 shows the dependence of shear stress on apparent shear rate for the control and compounds with coupling agent and compatibilizer, followed by regression results using the Ostwald-de-Waele (power law) model shown in table 3.

[GRAPH OMITTED]

Table 3 - regression results from figure 6 using [MATHEMATICAL
EXPRESSION NOT REPRODUCIBLE IN ASCII], where [Tau] is the shear stress
and [Gamma] is the apparent shear rate

Parameter        Control    0.50%        1%      0.5% MA-     1%
                           coupling   coupling     EPDM     MA-EPDM
                            agent      agent

K                 7.759     8.910       9.18       9.431      13.08
n                0.4539     0.340      0.9633     0.4125     0.3008
Regression
  coefficient,
  [R.sup.2]       0.999     0.997       1.00      0.998      0.995

It is seen that all compounds follow the power law model. The compound with 0.5% coupling agent gives the lowest shear stress over any shear rate range. This means that in extrusion, this compound will generate the lowest extruder head pressure at any given throughput.

Figure 7 shows the dependence of apparent viscosity on apparent shear rate for these compounds, as well as regression results using an exponential model. It is believed that the power law is valid at shear rates above 100 reciprocal seconds. The power law model failed to fit the curves and are not shown in figure 7. It is believed that the shear rate is outside of the range in which the power law is obeyed.

[GRAPH OMITTED]

Table 4 - regression results from figure 7 using [MATHEMATICAL
EXPRESSION NOT REPRODUCIBLE IN ASCII], where [Eta] is the apparent
viscosity, C, [Alpha] and [Beta] are constants and [Gamma] is the
apparent shear rate

Parameter     Control     0.50%     1% cou-   0.5% MA-   1% MA-
                         coupling    pling      EPDM      EPDM
                          agent      agent

C               15,363     9,197      9,681      5,315     1,564
[Alpha]       55,329.5    34,320     43,640     43,230    63,740
[Beta]            0.21    0.1219     0.1225     0.0779    0.1780
Reg. coef.,      0.996     0.999       1.00      0.998      1.00
  [R.sup.2]

Then the Carreau model (ref. 9) was applied in the form:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Where: [Lambda] is the characteristic time of the material [s]; n is the flow index and [[Eta].sub.0] the Newtonian viscosity. The Carreau model has a wider coverage for concentrated polymer solutions than power law model. When [Lambda] [much greater than] 1, Carreau model becomes power law mode.

The results are shown in figure 8. It is seen that only the compounds with coupling agent and with 0.5% MA-EPDM follow this model. The control and the compound with 1% MA-EPDM still appear to follow an exponential model.

[GRAPH OMITTED]

It is also interesting to note that the "n" values obtained from the Carreau model differentiates from those from the power law model except in the case of 0.5% coupling agent. The reason is not fully understood. It could be that the highly viscous compounds do not generate laminar flow, which is the basis for which the equations that describe apparent shear rate and viscosity are developed. As well, slip flow is commonly observed in highly filled rubber compounds. This means, in this study, slip flow may have occurred during the rheological measurement for those compounds. The non-slip-boundary condition, which is an essential assumption for the calculation of shear rate, is not necessarily met in this measurement. All these reasons cause the data discrepancy between power law and Carreau models.

Effect of magnetizable powder loading on compound magnetic force

In the characterization of magnetic materials, the B-H curve, or hysteresis loop, has been widely used due to its simplicity. This curve describes the cycling of a magnet in a closed circuit as it is brought to saturation, demagnetized, saturated in the opposite direction, and then demagnetized again under the influence of an external magnetic field. From this plot, the ramanent magnetization Mr can be calculated, which is proportional to the magnetic force of the material.

Table 5 - comparison of regressing results using
Carreau and exponential models

Parameter        Control      0.5%        1%       0.5%    1% MA-
Carreau model               coupling   coupling    MA-      EPDM
                             agent      agent      EPDM

[[Eta].sub.0]     36,590    355,000    677,100    41,320   48,500
[Lambda]          36,330      1.187      1.70     0.1305   18,400
n                 0.2932     0.3858     0.332       0      0.01832
Regression co-
  efficient,
  [R.sup.2]         0        0.998      0.996     0.996       0
Exponential
  model

C                 15,363                                    1,564
[Alpha]          55,329.5                                  63,740
[Beta]             0.21                                    0.1780
Regression
  coefficient,
  [R.sup.2]       0.996                                     1.00

It was believed that magnetic force was affected by the orientation of magnetizable particles, which is developed during compound processing and vulcanization. It was thought that applying a magnetic field to the compound before it is completely cured will align better particles and will possibly create particle chains that enhance the magnetic properties. A study was therefore conducted to verify this hypothesis by applying a magnetic field to uncured and cured compounds, as well as to compound that is cured in-situ in the magnetic field. The Mr values were calculated and compared. Figure 9 shows the B-H plot on the effect of state of cure of compound on its ramanet magnetization. The data are shown in table 6.

[GRAPH OMITTED]

Table 6 - ramanent magnetization values

Compound        Uncured    Cured before    In-situ cured
state of cure             applying field   in the field

Ramanent         0.115        0.127            0.122
magnetiza-
tion, Mr
(Tesla)

It is seen in table 6 that the cured compound gives the highest Mr value, followed by compound cured in-situ in the magnetic field. The uncured compound gives the lowest Mr value. The reason is not fully understood yet. More work is needed to clarify this aspect.

Results in table 6 also indicate that during magnetization, magnetizable particles do not move in the magnetic field to form particle-particle chains as previously expected. This is because the magnetic field is too weak to orient magnetizable particles inside the highly viscous compound.

Figure 10 shows the effect of magnetizable powder volume fraction on compound magnetic properties. Magnetic force correlates very well with Mr and it increases linearly with powder volume fraction. Higher attractive force is normally desired. However, in the current design proposal, a balance should be achieved between the attractive force (sealing force) and the effort for easy door opening.

[GRAPH OMITTED]

Design concept for magnetic seals

Figure 11 shows an example in the design of magnetic seals in automotive weatherstripping.

[ILLUSTRATION OMITTED]

Conclusions

* Magnetizable powder retards compound scorch time and curing, the higher the loading, the greater the retarding effect.

* Titanate coupling agent, at appropriate dosage, reduces compound viscosity and helps it retain mechanical properties after aging. However, the coupling agent retards compound scorch time and curing.

* MA-EPDM compatibilizer not only accelerates compound curing, but also enhances its mechanical properties. However, it increases compound viscosity and does not retain compound properties after aging.

* All the compounds follow the power law model for shear stress and apparent shear rate relationship. However, the apparent viscosity vs. apparent shear rate does not follow the power law model. Instead, those containing titanate coupling agent follow the Carreau model; others follow an exponential model.

* When magnetized, the cured compound gives the highest ramanent magnetization (Mr) value, followed by the compound cured in-situ within external magnetic field. The uncured compound gives the lowest Mr value. The magnetic field does not orient the magnetizable particles.

* The magnetic force increases linearly with magnetizable powder loading.

References

(1.) Y. Okamura. U.S. Patent 4,240,228 (1980).

(2.) A. Meria, U.S. Patent 4,700,509 (1987).

(3.) K. Gerdes, D. Corts, T. Jenkins, S. Lesmeister and L. Welle, U.S. Patent 4,644,698 (1987).

(4.) F. Mesnel, U.S. Patent 4,535,563 (1985).

(5.) W. White, U.S. Patent 4,192,101 (1980).

(6.) J. Keys and T. Gustafson, U.S. Patent 4,999,951 (1991).

(7.) P. Cittadini and A. Merla, U.S. Patent 5,257,791 (1993).

(8.) F.W.H. Kruger and W.J. McGill, J. Appl. Polym. Sci., 45, 563 (1992).

(9.) R.B. Bird, R. Armstrong and O. Hassager, Dynamics of Polymeric Liquids, Vol. I, p. 171.