Microstructure,physical,and mechanical properties of LDPE/UHMWPE blend foams

Abstract

An investigation was reported on the effect of foaming parameters on the microstructure, mechanical properties, and thermal conductivity of low-density polyethylene (LDPE) foams containing various amount of ultrahigh-molecular-weight-polyethylene (UHMWPE) as a reducer of chemical cross-linking. Azodicarbonamide (ADCA) and dicumyl peroxide (DCP) were used as foaming agent and cross-linking agent, respectively. The LDPE/UHMWPE blends were prepared in an internal mixer and foamed using a single-stage compression molding technique. Considering various parameters and their levels, optimization of Taguchi experimental design was carried out, an L₉ orthogonal standard array was selected and the efficient levels for different variables were calculated using analysis of variance (ANOVA) of the results. Also due to different objective functions investigated in this process, optimization of overall evaluation criteria (OEC) method was used. The results revealed that addition of UHMWPE leads to a significant increase in the storage modulus and complex viscosity of melt as well as a considerable decrease in gel content of blend foams compared to neat LDPE foam containing the same amount of DCP was observed. Also in presence of UHMWPE, the foam cell size was decreased compared to previous studies in the same condition. A linear relationship between relative density and thermal conductivity as well as cell size and thermal conductivity was observed. ANOVA results revealed that foaming temperature is the most effective parameter on foam properties and OEC results suggested 10 phr ADCA, 0.6 phr DCP, foaming temperature of 180°C, and 4 min soak time at foaming temperature are the optimum levels of parameters.

Introduction

The history of polymeric foams technology is traced from the late 1920s. Polymeric foams are cellular structure derived from expansion of blowing agent. Polymeric foams possess unique applications due to their excellent strength/weight ratio, superior thermal and sound insulation abilities, and energy-absorbing performance, which is governed by the polymer matrix, the cellular structure, and the gas composition. They are widely used in various fields including construction, transportation, packing textiles, sports applications, medical devices, and agriculture. In view of such broad application potential, the focus of research and development is now turning toward the production and optimization of foam products.

One important group of plastic foams is polyethylene (PE) foams, the foams produced using PE or blends of them play an important role in various plastic market. Ultrahigh-molecular-weight polyethylene (UHMWPE) is a kind of polyethylene with a molecular weight over 10⁶ g mol⁻¹. The long molecular chains of UHMWPE confer the following unique combination of properties are extensively used for various purposes. Most notable properties are high strength, excellent toughness, high resistance to chemicals, physical abrasion, and low friction coefficient. ¹ The processing of UHUWPE into foam has been of continuing interest for a number of years due to its high impact toughness, abrasion, and wear resistance.

In recent years, polymer blending technique is widely accepted as an economically practical way for the improvement of entirely polymeric materials, the scientific and industrial communities. Most commercial multicomponent polymer systems are two-phase blends that have advantages over the single-phase systems. It has been well documented that a number of physical properties and process ability of polyolefins can be improved by blending. ^2–4

UHMWPE has been widely used to optimize the property of polymers such as PE, polypropylene (PP), ethylene–propylene–diene monomer elastomer, polyaniline, and polyamide. ^5–9 Because of extremely high-molecular weight of UHMWPE, the melt viscosity is extremely high. This factor imposes a considerable hindrance on polymer processing, thus its application is limited to compression molding.

Matthews and Hoffman produced Low-density foam (0.02–0.09 g/cm³) from gels of UHMWPE with cell size of 5–75 μm. ¹⁰ Aydinli and Tincer introduced a new method to produce UHMWPE foams. ¹¹ Xin et al. studied the crystallization behavior and foaming properties of PP/UHMWPE, they showed that addition of UHMWPE led to decrease the mean cell size and narrowing cell size distribution. ¹² Rodrigue et al. investigated rheological behavior of UHMWPE foams; they found that increasing foaming agent led to decrease cell size because of competition between gas diffusion and coalescence. ^13,14

Previous researches showed that process conditions such as foaming time and temperature, blending, cross-linking level, and foaming agent content are parameters affecting physical and structural properties of PE foams mostly. ^15–19 Although using cross-linking agents such as DCP in PE foams lead to smaller cell size and better physical and mechanical properties, but it also causes nonrecyclable foams due to chemically cross-linking phenomena. Hence, in this research in order to increase the blend viscosity and decrease the cell size, UHMWPE, which is well known as a high viscose polymer, is used to decrease the chemical cross-linking agent content and increase the recyclability of produced foams. Also, this article will present some findings on the effect of different foaming parameters such as ADCA and DCP contents, foaming temperature, and foaming soak time on the morphology, mechanical properties, and thermal conductivity as well as the optimization of Taguchi experimental design is reported.

Experimental

Materials

Low-density polyethylene (LDPE) supplied by Bandar Imam petrochemical Co., Iran (melt flow index = 2 g min⁻¹ and $\left(\mathit{\rho }=0.92g\hspace{0.17em}{\mathrm{c}\mathrm{m}}^{-3}\right)$ and UHMWPE manufactured by Sigma Aldrich (Canada) (molecular weight = 4,500,000 and $\mathit{\rho }=0.94\hspace{0.17em}g\hspace{0.17em}{\mathrm{c}\mathrm{m}}^{-3}$ ) were used as base polymers. Azodicarbonamide (ADCA) provided by Merck (Germany) and dicumyl peroxide (DCP) supplied by Akzonobel (The Netherlands) were used as foaming agent and cross-linking agent, respectively.

Sample preparation

To prepare LDPE/UHMWPE blends with ADCA and DCP, a thermostatically controlled Brabender internal mixer was used. Considering 0.8 as the filling factor of the internal mixer, LDPE granules with UHMWPE powder was mixed at 190°C at 60 r min⁻¹ for 15 min. It should be noted that because of shear viscosity, temperature may rise up to 210°C. Previous researches stated that at this temperature, thermal degradation of LDPE and UHMWPE is negligible. After blending LDPE/UHMWPE, the formulation was removed from the mixer and cooled down to room temperature. In the second stage, the Brabender temperature was set at 120°C and granules of compounded polymer blends were loaded. ADCA was added, once torque became constant (after 3 min), and then DCP was added to the formulation after 2 min. Total time allowed for mixing in this stage, was 10 min. Finally, the blends were removed from the mixer and cooled down.

Foamed samples were prepared by compression molding (Brabender, Germany) via a single-stage foaming process. A specified weight of polymer blends was preheated in a cylindrical mold at 120°C. After applying 100 bar pressure, the temperature increased to 165°C and remained at this temperature for 5 min until the DCP decomposed. Then temperature increased to three different foaming temperatures and maintained for three different times based on the Taguchi experimental design. All these processes were performed based on previous studies.

Samples characterization

Oscillatory shear measurements in the linear viscoelastic region were performed on unfoamed blends, containing no DCP and ADCA, using a (Anton Paar) dynamic rheometer MCR 301 (Austria). Measurements were carried out in parallel-plate geometry (25 mm diameter and 1 mm thickness) at 180 and 200°C under nitrogen atmosphere. Frequency sweeps with an angular frequency velocity between 0.1 and 625 s⁻¹ were performed in the linear viscoelastic region at low strain of 1%. Samples were kept at zero shear rate for 5 min to equilibrate prior to measurements.

Foam density was measured according to the following equation:

${\mathit{\rho }}_{\mathrm{f}}=\frac{W_{\mathrm{f}}}{V},$

1

where W _f is the weight of foamed sample and V is the volume of sample. The relative density of the foam was then measured using equation (2) in accordance with ASTM-D3575 standard, as follows:

$\mathit{\rho }\mathrm{\ast }=\frac{{\mathit{\rho }}_{\mathrm{f}}}{{\mathit{\rho }}_{\mathrm{s}}},$

2

where ${\mathit{\rho }}^{\mathrm{\ast }}$ and ${\mathit{\rho }}_{\mathrm{s}}$ are the relative density and polymer matrix density of the foam, respectively. Also the foaming degree was calculated by:

$\mathrm{F}\mathrm{o}\mathrm{a}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{g}\hspace{0.17em}\mathrm{d}\mathrm{e}\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\hspace{0.17em}\left(\mathrm{\%}\right)=\left(1-\frac{{\mathit{\rho }}_{\mathrm{f}}}{{\mathit{\rho }}_{\mathrm{s}}}\right)\times 100.$

3

Cell microstructure of LDPE/UHMWPE foams was observed using a Tescan Vega II (Czech) scanning electron microscopy (SEM) operating at 15 kV. All samples were cut with a sharp blade and the cut surfaces were coated with a thin layer of gold according to ASTM-F1372 standard before SEM observations to enhance electrical conductivity. Analyzing of images was conducted by Image Pro. Express software, which enables determination of average cell size. Cell density of foams is defined as the number of cells per unit volume:

$N_0={\left(\frac{N{M}^2}{A}\right)}^{3/2}\left(\frac{1}{{\mathit{\rho }}^{\ast }}\right),$

4

where N ₀ and N are cell density and number of cells in the micrograph of area A in square centimeter with magnification of M.

The gel content of the samples was measured according to ASTM-D2765 standard. The foam samples of 300 ± 5 mg (initial weight) were put in a stainless steel 120 mesh cage and immerged in 350 ml of boiling xylene for 12 h. The remaining material (gel) was dried for 3 h at 150°C in a vacuum oven and weighted. The gel content was calculated using as follows:

$\mathrm{G}\mathrm{e}\mathrm{l}\hspace{0.17em}\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}\hspace{0.17em}\left(\mathrm{\%}\right)=\frac{\mathrm{G}\mathrm{e}\mathrm{l}\hspace{0.17em}\mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}}{\mathrm{I}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\hspace{0.17em}\mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}}\times 100.$

5

Compressive stress (σ) versus strain (ε) curves were measured using a Zwick/Roell testing machine (model T1-FR010, Germany) at room temperature. Three samples were tested for each specimen at a displacement rate of 5 mm min⁻¹. Samples were compressed to the strains of up to 80% between two parallel flat plates. These experiments were used to determine the foam Young’s modulus and stress in 10% strain based on ASTM-D1621 standard.

Thermal conductivity of the samples was measured using a homemade thermistor probe thermal conductometer designed and manufactured in Institute of Polymeric Materials, Sahand University of Technology. This instrument has a probe which is charged with energy pulse that increases the sample temperature. The time for temperature reduction was measured and thermal conductivity coefficient was calculated based on this time and temperature. Theoretical base of this instrument is calculating K (thermal conductivity coefficient) according to the following equation:

$T_{\mathrm{O}}\left(t\right)-T_{\mathrm{i}}=\frac{P(\mathit{\rho }{C}_P{)}^{0.5}}{4(\mathit{\pi }K{)}^{1.5}}\left[{\left(t-t_{\mathrm{P}}\right)}^{-0.5}-t^{-0.5}\right]$

6

where T _O is probe center temperature, T _i is probe primary temperature, P is electrical power, ρ is sample ensity, C _P is sample heat capacity, K is thermal conductivity coefficient, t is time, and t _P is energy pulse frequency. ²⁰

Taguchi experimental design

The method presented in this study is an experimental design process called the Taguchi experimental design method. Taguchi design, developed by Genichi Taguchi, is a set of methodologies by which the inherent variability of materials and manufacturing processes has been taken into account at the design stage. The application of this technique had become widespread in many US and European industries after the 1980s. The important of Taguchi design is that multiple factors can be considered at once and not only can controlled factors be considered but also noise factors. Although similar to design of experiment (DOE), the Taguchi design only conducts the balanced (orthogonal) experimental combinations, which makes the Taguchi design even more effective than a fractional factorial design. Using the Taguchi techniques, industries are able to greatly reduce product development cycle time for both design and production, therefore reducing costs and increasing profit. Moreover, Taguchi design allows looking into the variability caused by noise factors, which are usually ignored in the traditional DOE approach. ²¹

Since applying Taguchi experimental design requires the identification of factors affecting response quality characteristics, relevant literatures must be reviewed to screen the most important factors or conditions affecting PE foam properties. There are different factors involved in foam compression molding process. These factors are included in two main groups: the factors that could be altered during blending stage and the factors that could be altered during molding stage. Based on these facts and according to previous studies, experimental design has been carried out. Table 1 shows the control factors and their levels. The control factors are the basic controllable parameters used in the process, and noise factors are often uncontrollable variables in the process which may affect the response. The addition of noise factors in experimental design is optional. In this study, noise factors are neglected. To investigate the effect of UHMWPE content, experimental design was carried out in two stages. At the first stage, UHMWPE content was considered 5%wt and L₉ Taguchi standard orthogonal array was suggested. In the second stage for 10%wt UHMWPE, the same trend was carried out as well. Full factorial design requires 2³ × 2² = 32 experiments for each stage to clarify the effect of the parameters. With the selection of L₉ orthogonal array, the number of experiments required can be drastically reduced to nine. The standardized Taguchi-based experimental designs, L₉ orthogonal arrays, are shown in Tables 2 and 3. Data are analyzed using Qualitek-4 software.

Table 1.

Selected levels of foaming process parameters.

	Variable	Level
		Level 1	Level 2	Level 3
Blending stage	UHMWPE (%wt)	5	10	–
	ADCA (phr)a	5	10	–
	DCP (phr)a	0.3	0.6	–
Molding stage	Foaming temperature (°C)	180	190	200
	Foaming soak time (min)	2	4	6

UHMWPE: ultrahigh-molecular-weight polyethylene; ADCA: azodicarbonamide; DCP: dicumyl peroxide; LDPE: low-density polyethylene.

^aADCA and DCP content were selected based on LDPE/UHMWPE weight.

Table 2.

Designed L₉ Taguchi orthogonal standard array for samples containing 5%wt UHMWPE.

Exp.	UHMWPE (%wt)	ADCA (phr)	DCP (phr)	Foaming soak time (min)	Foaming temp. (°C)	Sample code
1	5	5	0.3	2	180	5UH-1
2	5	10	0.6	4	180	5UH-2
3	5	5	0.3	6	180	5UH-3
4	5	5	0.6	2	190	5UH-4
5	5	5	0.3	4	190	5UH-5
6	5	10	0.3	6	190	5UH-6
7	5	10	0.3	2	200	5UH-7
8	5	5	0.3	4	200	5UH-8
9	5	5	0.6	6	200	5UH-9

UHMWPE: ultrahigh-molecular-weight polyethylene; ADCA: azodicarbonamide; DCP: dicumyl peroxide.

Table 3.

Designed L₉ Taguchi orthogonal standard array for samples containing 10%wt UHMWPE.

Exp.	UHMWPE (%wt)	ADCA (phr)	DCP (phr)	Foaming temp. (°C)	Foaming soak time (min)	Sample code
1	10	5	0.3	180	2	10UH-1
2	10	10	0.6	180	4	10UH-2
3	10	5	0.3	180	6	10UH-3
4	10	5	0.6	190	2	10UH-4
5	10	5	0.3	190	4	10UH-5
6	10	10	0.3	190	6	10UH-6
7	10	10	0.3	200	2	10UH-7
8	10	5	0.3	200	4	10UH-8
9	10	5	0.6	200	6	10UH-9

UHMWPE: ultrahigh-molecular-weight polyethylene; ADCA: azodicarbonamide; DCP: dicumyl peroxide.

The signal-to-noise ratio (S/N) is simply a quality indicator by which the effect of a particular process parameter on the performance of the process or product is evaluated. In general, a better signal is obtained when the noise is smaller, so that a larger S/N ratio yields better final results. Depending upon the objective of quality characteristic there can be various types of S/N ratios. In order to maximize the response, the following S/N formulation is used:

$\frac{S}{N}=-10{\hspace{0.17em}\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\left[\frac{1}{n}\sum _{i=0}^n\frac{1}{x_i^2}\right].$

7

As well as when minimizing the response is considered, S/N ratio can be calculated as follows:

$\frac{S}{N}=-10{\hspace{0.17em}\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\left[\frac{1}{n}\sum _{i=0}^n{x}_i^2\right]$

8

where x_i is the value of quality characteristics and n is the number of replications. ²²

Analysis of variance (ANOVA) is a statistical technique to estimate the relative contribution of each control factors on the response (or objective function). The ANOVA can be conducted on the average or S/N data. The ANOVA based on the average data signify the factors which affect the average response rather than reducing the variation. However, ANOVA based on the S/N data takes into account both of these aspects, thus it was used in this study. Main equation to calculate ANOVA results are presented as below:

Correction Factor (C.F) is calculates by following equation:

$\mathrm{C}\mathrm{.}\mathrm{F}=\frac{{\left(\sum _{i=1}^n{y}_i^2\right)}^2}{N},$

9

where $\sum _{i=1}^n{y}_i^2$ is sum of the results from a DOE study and N is number of experiments. Sum of square for each factor is calculated as:

$S_{\mathrm{A}}=S_{A1}/N_{A1}+S_{A2}/N_{A2}+\dots -\mathrm{C}\mathrm{.}\mathrm{F},$

10

where for example N_A1 and A₁ are the total number of experiments and the total of result from experiments in which factor A is at level 1. Total sum of square is calculated as:

$S_{\mathrm{T}}=\sum _{i=1}^n{y}_i^2-C\cdot F$

11

and error sum of square is calculated as:

$S_{\mathrm{e}}=S_{\mathrm{T}}-\left(S_{\mathrm{A}}+S_{\mathrm{B}}+\dots \right).$

12

Degree of freedom (DOF) is calculated according to the following equation:

$f_{\mathrm{A}}=k-1$

13

$f_{\mathrm{T}}=N-1$

14

$f_{\mathrm{e}}=f_{\mathrm{T}}-f_{\mathrm{A}}-f_{\mathrm{B}}-\dots$

15

where f_A is the DOF of factor A, f _T is total DOF, f _e is error DOF, and k is number of levels for each factor. To calculate the variance, following equations shall be used:

$V_{\mathrm{A}}=\frac{S_{\mathrm{A}}}{f_{\mathrm{A}}}$

16

$V_{\mathrm{e}}=\frac{S_{\mathrm{e}}}{f_{\mathrm{e}}}$

17

Pure sum of square (S’) and contribution percentages (C.P) of each factor are calculated as:

$S_A^{\prime }=S_{\mathrm{A}}-\left(V_{\mathrm{e}}\times f_{\mathrm{A}}\right)$

18

$C\cdot P_{\mathrm{A}}=\frac{S_A^{\prime }}{S_{\mathrm{T}}}$

19

Since the optimum conditions for each individual response may differ, an overall evaluation criterion (OEC) is defined according to the Taguchi method, to obtain the best conditions at which all responses are in close proximity to their optimum values. For each criterion in OEC method, a relative weight is attributed in the range of 0–100%. The OEC value is calculated by the following equation:

$\mathrm{O}\mathrm{E}\mathrm{C}=\frac{X_1}{X_{1\mathrm{r}\mathrm{e}\mathrm{f}}}\times W{t}_1+\frac{X_2}{X_{2\mathrm{r}\mathrm{e}\mathrm{f}}}\times W{t}_2+\dots$

20

where X₁ is the evaluated value under criterion 1 and X_1ref is the highest value of X₁ and Wt₁ is the relative weight of criterion 1. ²³

Results and discussion

Rheological characteristics

It is well known that the viscosity of the constituent polymer melt blend has a significant influence on the final cell microstructure of the corresponding foams. Therefore, the rheological investigation were performed on the unfoamed blend samples, obtained in the first stage of Brabender internal mixer blending procedure, at two different temperatures (180 and 200°C) as well as at two different UHMWPE contents (5 and 10wt%). For comparison, the rheological behavior of LDPE at a temperature between 180°C and 200°C (190°C) was used in this study. As it can be seen in Figure 1, at both temperatures, storage modulus and complex viscosity increase with increasing UHMWPE content. Increasing molecular weight facilitates chain entanglement which makes difficult chains mobility. However, long chains lead to the higher molecular entanglements and increase the melt viscosity. Zhang and Rodrigue studied the rheological behavior of LDPE, HDPE, and UHMWPE at 180°C and reported similar results. ¹⁴

Figure 1.

LDPE/UHMWPE blends storage modulus and complex viscosity versus frequency at 180 and 200°C. UHMWPE: ultrahigh-molecular-weight polyethylene; LDPE: low-density polyethylene.

It is interesting to note that storage modulus and complex viscosity of sample containing 5% UHMWPE at 200°C is almost equal to that of neat LDPE at 190°C (Figure 1). This trend represents a very high drop of the melt viscosity of sample containing lower amount of UHMWPE at high temperature, while sample containing higher amount of UHMWPE at high temperature keeps its high viscosity.

As the efficiency of the melt mixture strongly depends on temperature, Figure 1 also demonstrates the storage modulus and complex viscosity at 180 and 200°C for samples containing 5 and 10% UHMWPE. It can be seen that storage modulus and complex viscosity decrease with increase of temperature, whereas, these reductions are more obvious at lower frequencies. Zhang and Rodrigue also studied the viscosity variation at different temperatures for neat UHMWPE; the results showed that with increasing temperature from 140°C up to 200°C, viscosity decrease from 10⁶ Pa·s to 10⁵ Pa·s. ¹⁴

Gel content

Polyethylene is difficult to dissolve in organic solvents at room temperature, but it tends to dissolve in xylene at an elevated temperature. It is important to note here that after the cross-linking is completed, the cross-linked gel is not dissolvable in the hot xylene. This, however, becomes a reliable method to check out the cross-linking percentage by measuring the portion of undissolved PE in the hot xylene. ²⁴ Many studies have investigated the effect of foaming and cross-linking agents on gel content. ^25–33 The results revealed that the control factor of gel content is mostly the foam cross-linking agent and the other parameters such as foaming temperature, foaming time, and foaming agent did not show significant influence on the gel content. ^31,34–36 The results suggested that increasing of cross-linking agent led to increase in the gel content. Based on the previous studies, increasing of the foaming agent slightly decrease the gel content. Marcilla et al. attributed this behavior to the reaction between residues of decomposed ADCA and DCP free radicals. ³² On the other hand, DCP half life at 165°C equals to 2.83 min. Therefore, regarding to the hot press time schedule, at foaming temperature, DCP is decomposed totally and the samples gel content do not depend on foaming temperature and foaming time. ³¹

Considering no influence of ADCA on gel content, the effect of introducing UHMWPE on the LDPE/UHMWPE blend foams gel content was investigated. The foams gel contents are listed in Table 4.

Table 4.

Gel contents of LDPE/UHMWPE foam samples with different UHMWPE and DCP contents.

Sample code	DCP (phr)	Gel content (%)
5UH-3	0.3	4.9
5UH-2	0.6	41.9
10UH-3	0.3	15.1
10UH-2	0.6	35.6

UHMWPE: ultrahigh-molecular-weight polyethylene; LDPE: low-density polyethylene; DCP: dicumyl peroxide.

In samples with the same UHMWPE content, gel content increases with increasing DCP content from 0.3 phr to 0.6 phr. This effect can be attributed to the cross-linking density increment which is conducted by decomposed DCP free radicals. Previous studies reported the same results. ^19,36 The results indicate that in samples with 0.3 phr DCP, increasing UHMWPE from 5%wt up to 10%wt, leads to the higher gel content. Very long chains of UHMWPE molecules have a barrier effect on the entrance of DCP molecules into UHMWPE phase. Therefore, with incorporating of UHMWPE to the LDPE/UHMWPE blends, LDPE phase content decreases and decomposition of cross-linking agent in LDPE phase leads to an increase in cross-linking density and gel content.

Davari et al. reported the gel content of neat LDPE containing 0.6 phr DCP and similar condition in the range of 46–50% foam. ³¹ Furthermore, at 0.6 phr DCP with increasing UHMWPE content from 5%wt up to 10%wt, the foams gel content decreases. It is obvious that gel content of LDPE/UHMWPE with the same DCP content is less than their achievement for neat LDPE foams and similar foaming process. This phenomenon can be related to the saturation of LDPE phase at 10% UHMWPE which can facilitate the radical termination reactions. On the other hand, high melt viscosity of LDPE/UHMWPE blends, which was revealed by rheological examinations, probably prevents penetration of radicals into the blends melt mixture and radical termination reactions increase.

Foam density and foaming degree

Foam density is an important parameter that defines relative ratio of solid and gas phases and is one of the most important parameters determining foam physical and mechanical properties. The results of density measurements are reported in Table 5. In Figure 2, the relative density of 5UH and 10UH foam samples is compared. Considering weight percentage of each component in LDPE/UHMWPE blends, solid polymer blend density is calculated as 921 kg m⁻³ for samples containing 5%wt UHMWPE and 922 kg m⁻³ for samples containing 10%wt UHMWPE.

Table 5.

Density of foam samples containing 5 and 10%wt UHMWPE prepared using L₉ Taguchi orthogonal array.

Sample code	Density (kg/m3)	Sample code	Density (kg/m3)
5UH-1	68.5 ± 0.7	10UH-1	78.5 ± 1
5UH-2	53.5 ± 0.7	10UH-2	58 ± 0.7
5UH-3	71.3 ± 0.6	10UH-3	84 ± 1.4
5UH-4	83 ± 1	10UH-4	86.5 ± 0.7
5UH-5	68 ± 0.7	10UH-5	76.3 ± 0.3
5UH-6	65.7 ± 0.8	10UH-6	75.5 ± 0.7
5UH-7	111.8 ± 1	10UH-7	119.5 ± 1.4
5UH-8	82.5 ± 0.5	10UH-8	68.5 ± 0.7
5UH-9	109 ± 2	10UH-9	81.7 ± 0.6

UHMWPE: ultrahigh-molecular-weight polyethylene.

Figure 2.

Relative density of 5UH and 10UH foam samples prepared using L₉ Taguchi orthogonal array.

The results indicate that at low and medium foaming temperatures (180 and 190°C), foam density of 5UH samples is lower than that of 10UH. For samples prepared at high temperature (200°C), contradictory trend is observed. Actually at low and medium temperatures, the effect of incorporating of UHMWPE to LDPE phase is more dominant. From rheological experiments, it is well known that the samples containing 10 wt% UHMWPE have higher viscosity compared to the samples with 5 wt% UHMWPE, which prevents bubble growth and therefore results in foam with higher density. On the other hand, due to the reduction of blends melt viscosity at higher temperature, the dominant mechanism is gas loss from the blends matrix which does not contribute in foam cells growth. Thus, the foam density in higher temperature (200°C) is higher than those in low and medium temperatures (180 and 190°C).

It was suggested that the temperature should produce a sufficiently low viscosity of the melt to allow cell formation and expansion as well as a sufficiently high viscosity of the melt to prevent cell collapse. At low temperature, viscosity of polymer melt and consequently melt strength is high and gas cannot escape of the melt. Increasing the foaming temperature, decreases the melt viscosity and gas loss phenomenon happens. This causes an increase in density. The gas loss phenomenon that occurs during foam processing could be correlated with the melt temperature. The diffusivity of the blowing agent at elevated temperatures was very high; therefore, if the processing temperature was too high, the gas could easily escape from the foam because of its higher diffusivity at elevated temperatures. In addition, as the cell expansion increased, the cell wall thickness decreased, and the resulting rate of gas diffusion between cells increased. Consequently, the rate of gas escape from the foam to the environment increased. Gas escape through the thin cell walls decreased the amount of gas available for the growth of cells, resulting in a lowered expansion and higher density. ³⁷ Furthermore, at low temperature the chain mobility decreases which leads to melt viscosity increment. Increasing melt viscosity prevents sufficient bubble growth that results in foam with higher density. Actually in polymers with high molecular weight and viscosity, produced gas from decomposed foaming agent is not able to overcome melt resistance which result in lower foam density compared to lower molecular weight ones. Generally, it is interesting to note that, there is a competition between the gas loss and viscosity resistance. Unexpectedly, sample 7 does not follow this trend. It can be attributed to short soak time of melt mixture at 200°C.

Figure 3 demonstrates comparison of foaming degree of samples containing 5 and 10%wt UHMWPE. It is noticeable that foaming degree of most samples are more than 90%. As expected, results of foaming degree indicate that for foam samples prepared at 180 and 190°C, foaming degrees of 5UH samples are higher than 10UH ones. While at 200°C due to higher melt viscosity of 10UH, which prevents gas loss, foaming degree is higher compared to 5UH samples. As well as in this case, an exception is observed for sample 7, as mentioned before; it can be related to the short soak time at 200°C. These results and the previous studies ³⁸ suggest that due to the competition between melt strength and gas loss and their effects on the expansion ratio, an optimized temperature is needed to achieve sufficient foam expansion.

Figure 3.

Foaming degree of 5UH and 10UH samples.

For a detailed investigation on the effect of control factors on the foam density, ANOVA was performed on the foam density results for 5UH samples (Table 6). These samples were chosen because of the better foam qualities that have been observed experimentally.

Table 6.

ANOVA results of density for 5UH foam samples.

Variable	DOF	Sum of square	Variance	Pure sum of square	Contribution percentage (%)
ADCA (phr)	1	0.944	0.944	0.03	0.089
DCP (phr)	1	0.085	0.085	0	0
Foaming temperature (°C)	2	24.171	12.085	22.343	65.135
Foaming soak time (min)	2	7.272	3.636	5.443	15.869
Error	2	1.828	0.913		18.907
Total	8	34.302			100

ANOVA: analysis of variance; DOF: degree of freedom; ADCA: azodicarbonamide; DCP: dicumyl peroxide.

ANOVA results emphasize that the effect of temperature on the foam density is more significant than the other factors. Soak time at foaming temperature and ADCA content are the other effective parameters, respectively. DCP content has no meaningful effect on the foam density. At a constant foaming time and temperature, Davari et al. reported ADCA as the effective parameter which influenced foam density. ³¹

Foam microstructure

SEM micrographs of different LDPE/UHMWPE foams are shown in Figures 4 and 5. These images illustrate that most the foam cells have pentagonal, dodecahedron, and tetracaidecahedra geometries. Almanza et al. also found out the similar results. ^39,40 It is shown that the cell walls are not completely smooth, even those are slightly twisted and wrinkled. As can be observed from these micrographs, cell wall shrinkage decreases with increasing of UHMWPE content. In addition to UHMWPE content, the other parameters such as foaming temperature and DCP content affect the shrinkage of cell wall. But detailed investigation of this subject needs more studies.

Figure 4.

SEM micrographs of different LDPE/5%wt UHMWPE blend foams at magnification ×100. SEM: scanning electron microscopy; UHMWPE: ultrahigh-molecular-weight polyethylene; LDPE: low-density polyethylene.

Figure 5.

SEM micrographs of different LDPE/10%wt UHMWPE blend foams at magnification ×100. SEM: scanning electron microscopy; UHMWPE: ultrahigh-molecular-weight polyethylene; LDPE: low-density polyethylene.

The average cell sizes of 5UH and 10UH foam samples are presented in Table 7. The minimum cell size of 48 μm is achieved which is significantly less than similar studies on neat LDPE foam at the same condition. ³¹ To eliminate the effect of foam density on the cell size, the normalized cell size was calculated as the cell size divided by relative foam density. Figure 6 compares the normalized cell size of 5UH and 10UH foam samples. As the results indicate, with increasing of UHMWPE content at low and medium temperatures (180 and 190°C), the average cell size increases. Increasing of UHMWPE content leads to decrease in LDPE phase in LDPE/UHMWPE blend. On the other hand, UHMWPE long chains don’t allow ADCA foaming agent to enter into the UHMWPE phase. Therefore, with increasing UHMWPE content, ADCA content increases in LDPE phase, which produces more decomposed foaming gas and results larger average cell size. At high temperature (200°C), LDPE/UHMWPE blends viscosity drop off, consequently with increasing UHMWPE content in blends, the foam cell sizes decrease. Particularly it is noticeable that in these foam samples cross-linking agent content is lower than the other samples. In this case, UHMWPE has important role, and with increasing UHMWPE, melt strength increases which has barrier effect on cells growth and coalescence that results in smaller cell size.

Table 7.

Average cell size of different foam samples containing 5 and 10%wt UHMWPE.

Sample code	Cell size (μm)	Sample code	Cell size (μm)
5UH-1	60 ± 1.7	10UH-1	230.5 ± 1.2
5UH-2	48 ± 0.9	10UH-2	86 ± 1.1
5UH-3	85 ± 1.4	10UH-3	96 ± 2.6
5UH-4	58.5 ± 0.7	10UH-4	74 ± 0.2
5UH-5	70 ± 2.6	10UH-5	84 ± 1.2
5UH-6	103 ± 4.9	10UH-6	155 ± 6.2
5UH-7	268 ± 7.4	10UH-7	367 ± 11.5
5UH-8	120.5 ± 1.1	10UH-8	85 ± 3.4
5UH-9	181 ± 2.3	10UH-9	97 ± 7.5

UHMWPE: ultrahigh-molecular-weight polyethylene.

Figure 6.

Comparison between normalized cell size of different 5UH and 10UH foam samples.

Sample 7 was found to be in contrast with samples 8 and 9; in this sample, the average cell size of 5UH is less than 10UH one. This behavior could be attributed to the short soak time at foaming temperature as mentioned in previous section. In such short soak time (2 min), the viscosity drop at low and medium temperature is not high enough; therefore with increasing soak time at high enough temperature, the viscosity drop is more considerable.

Although it is expected that the larger average cell size leads to the lower foam density, the results show contradictory trend. For example at 180°C; however, the foam density of 5UH is less than 10UH one, its cell size is smaller than that. Rodrigue et al. observed similar trend for neat UHMWPE foams. ¹³ In those samples dissimilarly to LDPE and HDPE foams, increasing ADCA content resulted in smaller cell size. This effect could be related to competition between gas loss and cells coalescence.

Figure 7 shows the cell density of 5UH and 10UH foam samples. At foaming temperatures 180 and 190°C, the cell densities decrease with increasing of UHMWPE content. At temperature 200°C, opposite trend is observed. In samples 5UH-7 and 10UH-7, the cell densities are lower compared to the other samples.

Figure 7.

Comparison between cell density of 5UH and 10UH foam samples.

According to better foam quality of 5UH samples, the results of normalized cell size and cell density of these samples have been analyzed in this study. S/N ratio was calculated with considering the smaller the better cell size and larger the better cell density criteria. In Figure 8, S/N graphs of cell size for all efficient parameters are indicated. As the results show, with foaming agent at lower level (5%), DCP cross-linking agent at lower level (0.3%), foaming temperature at second level (190°C), and soak time at foaming temperature at third level (6 min), minimum cell size is achieved.

Figure 8.

S/N variation graphs for cell size of 5UH foam samples. S/N: signal-to-noise ratio.

ANOVA results of cell size for 5UH foam samples are presented in Table 8. These results represent that foaming temperature is the most effective parameter. The foaming and cross-linking agents have moderate influence on the average cell size, respectively, and soak time at foaming temperature has no significant effect.

Table 8.

ANOVA result for average cell size of 5UH foam samples.

Variable	DOF	Sum of square	Variance	Pure sum of square	Contribution percentage (%)
ADCA (phr)	1	7.47	7.47	3.449	1.78
DCP (phr)	1	9.742	9.742	5.72	2.953
Foaming Temperature (°C)	2	44.712	72.356	136.669	70.563
Foaming Soak Time (min)	2	23.713	11.856	15.67	8.09
Error	2	8.043	4.021		16.614
Total	8	193.683			100

ANOVA: analysis of variance; DOF: degree of freedom; ADCA: azodicarbonamide; DCP: dicumyl peroxide.

Figure 9 illustrates S/N graph of cell density for all control parameters. According to S/N results, 5% of foaming agent, 0.6% of cross-linking agent, foaming temperature of 180°C, and soak time at foaming temperature of 4 min lead to the highest cell density. ANOVA results of cell density are presented in Table 9 and indicate that foaming temperature with contribution percentage of 69.79% is the most effective parameter influencing cell density, and after that, the foaming time, DCP, and ADCA contents are subsequent effective parameters, respectively.

Figure 9.

S/N variation graphs for cell density of 5UH foam samples. S/N: signal-to-noise ratio.

Table 9.

ANOVA results for cell density of 5UH foam samples.

Variable	DOF	Sum of square	Variance	Pure sum of square	Contribution percentage (%)
ADCA (phr)	1	69.661	69.661	16.744	0.733
DCP (phr)	1	99.567	99.567	46.65	2.044
Foaming Temperature (°C)	2	1698.499	849.249	1592.665	69.79
Foaming soak time (min)	2	308.516	154.258	202.682	8.881
Error	2	105.833	52.916		18.522
Total	8	2282.078			100

ANOVA: analysis of variance; DOF: degree of freedom; ADCA: azodicarbonamide; DCP: dicumyl peroxide.

Figure 10 presents comparison between 5UH and 10UH sample’s cell size distribution. Obviously, at 180 and 190°C, 5UH foam samples have narrower cell size distribution than 10UH samples. Higher LDPE phase in 5UH foam samples compared to 10UH ones leads to the lower ADCA content in LDPE/UHMWPE blend foams that reduces the possibility of cell destruction in foams, which generate smaller cells that result in narrower cell size distribution. Dissimilarly at 200°C with increasing UHMWPE content in foam blends, the melt strength increases, which prevent bubble growth and coalescence and result in narrower cell size distribution. On the other hand, at 180°C with increasing of foaming time, the cell size distribution of 5UH and 10UH samples are almost the same. At 200°C, with increasing of foaming time, distribution curves are different. This trend can be attributed to decrease of melt strength with time at a specific temperature. Furthermore, it is shown that at a constant foaming time with foaming temperature increment, 10UH sample’s cell size distribution are narrower than 5UH ones.

Figure 10.

Cell size distribution of different foam samples.

Mechanical properties of LDPE/UHMWPE foams

Figure 11 demonstrates the stress-strain curves of 5UH and 10UH foam samples. Based on the samples response to compression force, it is clear that LDPE/UHMWPE blend foams are classified as elastic foams, because there is no stress reduction representing yield point between linear and plateau regions.

Figure 11.

Stress-strain behaviors of different LDPE/UHMWPE foam samples. UHMWPE: ultrahigh-molecular-weight polyethylene; LDPE: low-density polyethylene.

The stress–strain results show that in samples 1 to 5 decreasing UHMWPE content in blends leads to increase of foam modulus. This result could be related to smaller cell size and narrower cell size distribution in samples containing lower UHMWPE content. In narrow cell size distribution, stress faces with small cells with similar average cell size, so the peak stress is higher, while in broad cell size distribution, the stress primarily concentrates on larger cells and afterward, distributes on smaller cells. Due to the gas inside the cells, contribution of internal pressure for larger cells is low due to the higher probability of breaking for the more strained cell walls. In other words, higher external pressure required to overcome the internal gas pressure for smaller cells. ⁴¹ On the other hand, it is shown that with reducing UHMWPE content in the blend foams, area under the stress–strain curve increases, which demonstrates increasing of the samples toughness. Furthermore, with decreasing UHMWPE content, cell density increases which leads to increase in cells surface area and results in more energy absorption and starting densification region at higher stresses.

In sample 6, 5UH and 10UH samples have similar trend till 25% strain. This similarity can be attributed to competition between foam density, cell density, and cell size distribution. In sample 7, it is clear that at linear region the trends of stress–strain variation in both samples are similar. With increasing UHMWPE content in samples 8 and 9, cell size decreases and a narrower cell size distribution is observed, so modulus increases. Moreover, for both samples with increasing UHMWPE content in blends, cell density increases, which results in higher densification stress.

The results of stress at 10% strain, elastic modulus of samples containing 5 and 10%wt UHMWPE are presented in Table 10, and Figure 12 represents the normalized stress at 10% strain respect to foam density of these samples. These results suggest that at low and medium foaming temperature with increasing of UHMWPE content in the blends, the average cell size increases and therefore stress decreases. At high temperatures, increasing of UHMWPE content leads to decrease in cell size and increase of stress.

Table 10.

Stress at 10% strain and elastic modulus of foam samples containing different UHMWPE.

Sample	Stress in 10% strain (kPa)	Elastic modulus (kPa)	Sample	Stress in 10% strain (kPa)	Elastic modulus (kPa)
5UH-1	117.32 ± 2.4	10.8 ± 0.6	10UH-1	114.9 ± 0.2	12.4 ± 0.6
5UH-2	138.5 ± 1.2	13.7 ± 0.8	10UH-2	121 ± 0.8	14.4 ± 0.2
5UH-3	143.1 ± 3.2	16.3 ± 1.7	10UH-3	77 ± 2.8	10.7 ± 0.7
5UH-4	195.7 ± 0.1	22.8 ± 3.4	10UH-4	122.1 ± 3.1	15.5 ± 0.1
5UH-5	216.1 ± 3.8	23.6 ± 1.3	10UH-5	83.4 ± 1.3	10.6 ± 0.3
5UH-6	132.5 ± 2.4	12.4 ± 1.4	10UH-6	131.8 ± 0.4	18.1 ± 1.1
5UH-7	121.6 ± 1	14.2 ± 2.5	10UH-7	119.6 ± 2.5	11.2 ± 0.1
5UH-8	77.2 ± 0.4	11 ± 1.8	10UH-8	100.5 ± 1.4	9.4 ± 0.9
5UH-9	101.8 ± 1.1	13.5 ± 5.9	10UH-9	262.3 ± 2.9	24.4 ± 1.5

UHMWPE: ultrahigh-molecular-weight polyethylene.

Figure 12.

Normalized stress in 10% strain respect to foam density for different foam samples.

Figure 13 shows variation of compressive stress at 10% strain versus foam density of 5UH samples. As it is shown, increase of density does not have significant effect on compression stress at 10% strain. It is remarkable that stress at 10% strain for the majority of foams is observed at the first stress–strain region or at the beginning of the second stress-strain region. This result indicates that for foams used in elastic region, density variation has no significant effect on mechanical properties and foam density control is not too much essential. ⁴²

Figure 13.

Variation of stress at 10% strain versus relative density for 5UH foam samples.

Figure 14 shows normalized compressive stress versus normalized cell size. This diagram does not follow a notable trend but generally, increase of cell size leads to decrease of compressive stress. With increasing cell size, gas phase increase compared to polymer phase. Since the stress is held by polymer struts and walls in the foam cells, thus, reducing amount of polymer phase leads to decrease of compressive stress.

Figure 14.

Variation of normalized compressive stress versus normalized cell size for 5UH foam samples.

To determine the optimal parameters affecting the compressive stress at 10% strain of the samples due to scattered data of different replications, mean values statistical analysis was used instead of S/N analysis. Considering bigger the better quality characteristic, mean value diagrams of compressive stress for all efficient parameters of 5UH samples are plotted in Figure 15. The results reveal that with 5 phr of foaming agent, 6 phr of cross-linking agent, foaming temperature of 190°C, and soak time of 6 min, maximum stress is achieved. In Table 11, ANOVA results of mean value data for 5UH samples are presented. It is obvious that cross-linking agent is the most efficient parameter to determine compressive stress and then, foaming agent and foaming temperature are important, respectively. The foaming soak time has no influence on compressive stress.

Figure 15.

Mean value variations of compressive stress in 10% strain for 5UH foam samples.

Table 11.

ANOVA results of mean value for compressive stress of 5UH foam samples.

Variable	DOF	Sum of square	Variance	Pure sum of square	Contribution percentage (%)
ADCA (phr)	1	10,611.67	10,611.671	9073.721	17.723
DCP (phr)	1	17,038.413	17,038.413	1550.463	30.276
Foaming Temperature (°C)	2	4778.355	2389.177	1702.455	3.325
Foaming soak time (min)	2	1850.649	925.324	0	0
Error	2	16,917.448	1537.949		48.676
Total	8	51,196.539			100

ANOVA: analysis of variance; DOF: degree of freedom; ADCA: azodicarbonamide; DCP: dicumyl peroxide.

Thermal conductivity

It is well known that the thermal conductivity of foams is due to four different mechanisms: conduction along the cell walls and struts of the solid polymer (λ _s), conduction through the gas (λ _g), thermal radiation (λ _r), and convection within the cells (λ _c). The total heat transfer can be predicted as the sum of the heat transfer by the four mechanisms considered separately. ²⁷ Figure 16 shows thermal conductivity coefficient of 5UH foam samples. It is obvious that thermal conductivity coefficient of foams prepared at 200°C is higher than that at 180 and 190°C. At 200°C due to the gas loss from polymer matrix, high density is achieved, and on the other hand, because of the lower viscosity of melt mixture, the foam cell sizes are larger and the cell wall surfaces are increases. Therefore, the contribution of thermal conductivity through solid phase and the thermal radiation increase because the cell walls are more transparent than the struts. ⁴¹ To eliminate the effect of foam density, thermal conductivity coefficients of the samples are normalized by dividing it to the relative density $\left({\mathit{\rho }}_f/\mathit{\rho }\right)$ of each sample and the results are presented in Table 12.

Figure 16.

Thermal conductivity coefficient of 5UH foam samples.

Table 12.

Normalized thermal conductivity coefficient of 5UH foam samples.

Sample code	Normalized thermal conductivity (mW mK−1)
5UH-1	427.2
5UH-2	395.6
5UH-3	411
5UH-4	396.5
5UH-5	492.5
5UH-6	361.4
5UH-7	556.2
5UH-8	542.5
5UH-9	464.2

According to the previous studies, foam density and cell size are the main factors in foam thermal conductivity. ^34,43–45 Figure 17 illustrates the thermal conductivity coefficient of 5UH foam samples versus foam density. As the results indicate, the thermal conductivity increases linearly with foam relative density. Previous research showed that in neat LDPE foam that have foam density more than 50 kg m⁻³, with increasing foam density, their thermal conductivity was increased. It could be related to increase of solid phase in cell walls and consequently increase of heat transfer through the solid phase. ⁴⁶

Figure 17.

Thermal conductivity coefficient variation versus relative density of 5UH foam samples.

Thermal conductivity variation versus cell size is illustrated in Figure 18. In order to eliminate the effect of foam density on the results, normalized thermal conductivity coefficient and normalized cell size were used. It can be observed that with increasing of cell size, thermal conductivity coefficient of foams increases. Increasing cell size, increases transparent cell walls surface area which causes less barrier against radiation, so thermal conductivity increases. ⁴⁶

Figure 18.

Normalized thermal conductivity coefficient versus normalized cell size of 5UH foam samples.

Figure 19 illustrate S/N graphs of thermal conductivity coefficient. As it can be seen, optimum contents of foaming agent, cross-linking agent, foaming temperature, and foaming soak time are 5 phr, 0.6 phr, 180°C, and 4 min, respectively. ANOVA results are presented in Table 13. According to the variables’ contribution percentage, the foaming temperature, foaming soak time, and foaming agent content are the effective parameters on thermal conductivity, respectively. In this case, cross-linking agent has no significant effect.

Figure 19.

S/N graphs of thermal conductivity for 5UH foam samples. S/N: signal-to-noise ratio.

Table 13.

ANOVA results of thermal conductivity for 5UH foam samples.

Variable	DOF	Sum of square	Variance	Pure sum of square	Contribution percentage (%)
ADCA (phr)	1	4.568	4.568	2.853	7.519
DCP (phr)	1	0.75	0.75	0	0
Foaming Temperature (°C)	2	19.615	9.807	16.185	42.652
Foaming soak time (min)	2	9.852	4.791	6.152	16.213
Error	2	3.43	1.715		33.616
Total	8	37.947			100

ANOVA: analysis of variance; ADCA: azodicarbonamide; DCP: dicumyl peroxide; DOF: degree of freedom.

Overall evaluation criteria

As it was mentioned, due to occurrence of different optimum conditions for each objective function, OEC method is suggested to achieve a unique condition. Foam density, cell size, and thermal conductivity with quality characteristic of smaller the better; cell density and stress in 10% strain with quality characteristic of bigger the better were chosen as OEC responses. Due to discrepancy of replicated data, analysis of the mean values was used instead of S/N analysis. Considering the same weight for all objective functions (20% relative weight for each response), the mean value results of OEC were calculated and illustrated in Figure 20. OEC results indicate that with considering 10 phr of foaming agent (ADCA), 0.6 phr of cross-linking agent (DCP), 180°C of foaming temperature, and 4 min of soak time at foaming temperature; minimum foam density, minimum cell size, maximum cell density, maximum compressive stress at 10% strain, and minimum thermal conductivity coefficient are achieved. As ANOVA results (Table 14) show, efficient parameters are foaming temperature, foaming soak time, and DCP content, respectively. In this case, ADCA content has no significant effect on OEC results.

Figure 20.

OEC mean value variations of 5UH foam samples. OEC: overall evaluation criteria.

Table 14.

ANOVA results of OEC mean value for 5UH foam samples.

Variable	DOF	Sum of square	Variance	Pure sum of square	Contribution percentage (%)
ADCA (phr)	1	1.049	1.049	0	0
DCP (phr)	1	173.318	173.318	158.253	1.371
Foaming temperature (°C)	2	9558.155	4779.077	9528.025	82.75
Foaming soak time (min)	2	1640.353	820.176	1610.223	13.955
Error	2	165.714	15.064		2.099
Total	8	11,538.591			100

ANOVA: analysis of variance; OEC: overall evaluation criteria; ADCA: azodicarbonamide; DCP: dicumyl peroxide; DOF: degree of freedom.

Conclusion

The main purpose of this research is to study the effect of various foaming parameters on the microstructure, mechanical properties, and thermal conductivity of LDPE/UHMWPE blend foams prepared by compression molding method with DCP and ADCA as cross-linking and foaming agents, respectively. Furthermore, to investigate the effect of process parameters on the properties of LDPE/UHMWPE blends foams and to find the optimum conditions of parameters, the Taguchi experimental design method was used. With considering two levels of UHMWPE content (5 and 10%wt.), the process parameters including DCP and ADCA contents both in two levels and foaming temperature and foaming time both in three levels were selected. A Taguchi L₉ orthogonal standard array was used for experimental design and the ANOVA method was considered for statistical analysis. The results showed that adding UHMWPE increased the storage modulus and complex viscosity strongly. Moreover, compared to previous researches, gel content of LDPE/UHMWPE was less than gel content of neat LDPE containing 0.6 phr DCP. Results also indicated that at low and high foaming temperature, melt strength, and gas loss are the overcoming phenomena, respectively. This caused increasing relative density and normalized cell size and decreasing cell density and normalized compression stress with increasing UHMWPE content at low and medium foaming temperature (180 and 190°C), whereas at high foaming temperature, the procedure is contrariwise. Results of 5UH samples revealed that increase of density does not have significant effect on compression stress at 10% strain but increases thermal conductivity coefficient. Also increase of cell size decrease compressive stress in 10% strain and increase thermal conductivity of 5UH foam samples. According to the OEC results, to obtain the least relative density, cell size and thermal conductivity and the most cell density and compression stress, optimum levels for the four factors including foaming agent content, cross-linking agent content, foaming temperature, and foaming time should be 10 phr, 0.6 phr, 180°C, and 4 min, respectively. Also the ANOVA tables introduced the foaming temperature as the most effective parameter.