Factors Affecting Fatigue in H R Adduct Foam - Statistically Oriented Preparation and Evaluation* H. W. Wolfe, Jr., D. F. Brizzolara, and J. D.

Byam

E. I. duPontdeNemours & Co. Elastomers Laboratory

Wilmington, Delaware

Introduction

H igh

isocyanate

moldability.

resilience

(HR) foam, since

its introduccold cure foam, has changes both in quality and

tion about six years ago

as

undergone numerous application. Initially, the greatest acceptance and ultimate use was in automotive seating. This acceptance was based on (1) ease of molding with the potential of a lower reject rate, (2) overall esthetics, (3) improved surface porosity and fewer closed cells which minimized the &dquo;pneumatic&dquo; feel, and (4) improved comfort due to high resilience and sag factor. Other foam properties, as matched against the application requirements, were considered more than adequate. Since the introduction of the first cold cure foam (high resilience foam) based on 100% crude MDI and a high molecular weight primary triol numerous HR foam systems have been developed. Basically, each one of these systems incorporated a highly reactive primary triol, a hardness building reagent, and a gelation extending reagent as shown in Table 1. The polyol is generally a conventional PPG type capped with ethylene oxide, having a molecula.r weight of 4500 to 6500. Hardness is developed by means of graft polyols, low molecular weightt diamines or triols, isocyanates such as undistilled TDI or combinations of more than one of these. Gelation is achieved by means of the same diamines or triols plus osocyanates such as polymeric MDI, undistilled TDI, prepolymers, or combinations of these various reagents.

Adduct H R Foams More recently a new isocyanate, identified as adduct isocyanate, has been shown to provide another means of preparing HR foams. Foams based ’~This paper was originally presented at the SPI Cellular Plastics in Transportation Conference on October 6, 7, and 8, 1975. As a Conference Paper, it has not been subjected to the Journal’s normal review procedures, but is reprinted here as a service to our subscribers.

48

this

have excellent processibility and Gelation is achieved by means of the isocyanate and does not require another reagent to on

assist in the processing. Likewise, much of the hardness is obtained from the isocyanate rather than from graft polyols or low molecular weight diamines or triols, although these reagents can be used to achieve additional foam hardness when desired. Table 1 shows the simplicity of an HR foam formlation based on adduct isocyanate. Tlus formulation can also be modified by adding reagents such as graft polyol and crosslinking agents such as metylene dianiline or triethanol amine.

What is an Adduct

Isocyar~ate?

isocyanate discussed in this paper is the reaction product of a polyamine with excess HYLENE@ TM-65 ~ . This material (Table 2)), identified as HLG-3897, is a straw colored liquid having a viscosity of 25-30 cps at 25°C, an amine equivalent of about 106, and an average functionality of 2.5. Because of the physical similarity to TDI, the adduct can be processed with ease in any equipment normally used for metering and mixing one-shot urethane foams. Due to the presence of an excess of TDI in the adduct isocyanate, handling precautions should be identical to those used for TDI. The adduct

1 HYLENE® TM-65 is a distilled isomer blend of 65% 2,4 and 35% 2,6 toluene diisocyanate.

third

cycle at both 25% and 65% is low.

Polyols the isocyanate are the two major HR foam system not only because they constitute the major portion of the system, but because they contain the functional groups through which the urethane polymer is formed. Since the basis of the Du Pont system is adduct isocyanate, this component was held constant throughout this study. In addition, a catalyst system was chosen which would provide adequate processing for the range of polyols to be studied. Surfactant type and level were also held constant. Based on these ground rules, a systematic, statistically designed study of the effect of polyol on fatigue in adduct isocyanate HR foam was undertaken. The formulation used for most of the polyol The

polyol and

components in

Scope of Study As the demand for HR foam increases in current new applications, so does the demand for improvements in properties and performance to meet new requirements. One of the demands more recently created is for improvement in fatigue resistance as applied to automotive seating. Thus, the main objective of this paper is fatigue resistance in HR foams, specifically HR foams based on the adduct isocyanate. Other foam properties were measured but these were secondary in import-

and

experiments

is

an

given in Table 3.

ance.

Fatigue Testing The subject of

complex enough

fatigue

to warrant

measurement is, itself, a study of its own. A

natural concern for any type of measurement chosen is the degree to which the results of a given test can be correlated to actual performance of a foam in applications such as mattresses, furniture and automotive cushions. Du Pont has, in the past, compared the static fatigue test to in use performance of ( 1 ) hotel mattresses for 90 days, (2) topper pads in taxicabs, and (3) furniture cushions in a YMCA lobby for 250 days. In all three comparisons, excellent correlation was obtained. Basically, the static fatigue test is a technique of measuring stress relaxation. Molded foam blocks measuring 15 X 15 X 4Y2&dquo; are conditioned at 72°F and 50% R.H. and then measured for ILD at both 25% and 65%. The foams are placed in clamps and compressed to 75% deflection. After being held under compression for 17 hours at 75°F and 50% R.H., the foams are released from the clamps. Precisely 5 minutes after release from the clamps, ILD’s at 25 and 65% are again determined. The loss of ILD at 25 and 65% deflections is calculated for each 17 hour period under compression, using the original ILD as the base point. A total of 3 cycles are determined on each foam measured. Generally, the ILD loss (fatigue) at 25% is greater than at 65%. Good fatigue resistance is generally shown by three conditions being met: 1. Three cycle ILD loss at both 25% and 65% is low in absolute terms. 2. Three cycle ILD loss at 25% is close or equal to the three cycle ILD loss at 65%. 3. The change in ILD loss between the first and

The this

principal objectives

of the

of

polyol phase

study were:

1. ascertain whether polyols have an effect on fatigue resistance in an adduct isocyanate HR foam system ; 2. attempt to identify those variables of the polyols studied which determine their effect on fatigue resistance; 3. select an optimum polyol formulation on which to base the remainder of this study. A few conditions were placed on the selection polyols for this study. First, they had to be commercially available. Second, they had to, as a group, embody a number of structural variations; e.g., molecular weight, functionality, primary vs secondary hydroxyl conventional vs graft, so that some trends could be determined regarding the second objective. A description of the polyols chosen for this study is given in Table 4. Note that this list includes

of

49

from a number of different sources. Althis possibly interjects some unknown variables into the study, the basic statistical trends are valid. A pure study of the effect of the various structural features of polyols on fatigue would require a series of specially prepared polyols. Such a series is not commercially available and, therefore, beyond the scope of this study.

polyols though

Figure 2. Polyol Mixtures Nominal Functionality Contours.

tionalities

might be different for some points, especially at higher molecular weight, the relative ordering of polyols is valid. The influence of polyol variation on fatigue is summarized in Figures 3, 4, 5 and 6. There is an

The outlined

specific experimental points

as originally given in Table 5. As experimentation in this phase progressed, additions to and deletions are

from this list

were

polyols

studied, they

made. Note that where blends of were constructed to give constant equivalent weight within a given series. Figures 1 and 2 show graphically the nominal molecular weight and functionality contours resulting from the various polyol blends considered. The were

Figure 3. Static Fatigue

25% I LD Loss - 1 Cycle (%).

Figure 1. Polyol Mixtures Nominal Molecular Weight Contours.

nominal molecular weights and functionalities were used in drawing these contours in order to put all polyols on the same basis, since exact functionalities are now known expecially for high molecular weight &dquo;triols&dquo;. In this connection, it is reiterated that in this study the objectives were to determine if polyol affects fatigue, the trends of this effect, and the selection of the optimum system within the scope of the study. Therefore, although the actual func-

50

Figure

4. Static

Fatigue 65% I LD Loss -

1

Cycle (%).

unmistakable trend in all cases toward improved fatigue resistance as polyol composition approaches the vicinity of Point I on these graphs. What does this tell us in terms of polyol structure and our objectives? Comparing the ILD loss contours (Figure 3-6) and the functionality and molecular weight contours (Figures 1 & 2), leads to the con-

51

importance of these variables fatigue. Therefore, in terms of the stated objectives of this phase of the study; . polyols do affect fatigue in adduct isocyanate HR foam, . polyol molecular weight, functionality, and chemical structure affect fatigue resistance, . the specific &dquo;6500 MW, primary triol&dquo; used in this study gives the best fatigue resistance and is regarding the they relate

as

relative to

the basis for the remainder of this work.

Polyol variation influences other properties as as fatigue. A major effect is seen in foam hard-

well

measured by either 25% or 65% ILD, or 50% C/D, the trend being that lower molecular weight ness as

Figure

5. Static Fatigue 25% LD Loss - 3 Cycle (%).

polyols produce harder foams. Another property significantly affected by polyol composition strength which decreases as polyol molecular

that is is tear

weight increases. Other properties are affected by polyol composition, but the trends are more complex and values generally remain within acceptable limits

over

the whole range.

Catalysts Having established an optimum polyol with the isocyanate HR system, emphasis was shifted to a study of the effect of catalyst on fatigue. Four catalysts were selected for study on the basis of their commercial availability and their principal adduct

Figure

6. Static

Fatigue

65°la I LD Loss - 3

Cycle (%).

clusion that improvement in fatigue occurs as molecular weight and functionality are increased. Although this conclusion is believed to be generally true, lack of information regarding actual functionality, especially for the high molecular weight triols, leads to some uncertainty. Other factors which are

unknown, for example, percent primary hydroxyls in the

and details of manufacture another also contribute to some measure of nncertainty with respect to absolute conclusions; nevertheless, the trends noted are statistically valid for the group of polyols studied. It is, therefore, concluded that:

from

~

~

capped polyols,

one source to

polyol represented by Point 1 on the graph gives the best fatigue resistance, dilution of this polyol with any lower molecular weight diol or triol, primary or secondary, conventional or graft leads to reduced fatigue resistthe

ance, ~

trend toward

improved fatigue resistance ocpolyol mixture molecular weight and functionality are increased, and that these variables of molecular weight, functionality, and chemical structure are not independent on one a

curs

as

another.

What this

52

study

does

not

provide is information

function. These and the formulations used in the catalyst study are shown in Tables 6 and 7. As with the polyols, this study was approached statistically. After a few experiments, it became obvious that fatigue resistance is improved by reducing or eliminating dibutyl tin dilaurate (DBTDL). The revised catalyst design is shown in Figure 7.

,

Figure

7.

Generally, the catalyst system have at least

Figure 9. Catalyst Study - Static Fatigue 65% I LD Loss - 1 Cycle (%).

Catalyst Study - Experimental Design.

as

significant

an

effect

was

on

found

fatigue

as

’

to

the

in the adduct isocyanate HR system, in which the basic building blocks are put together is as important as the components themselves. The effect of catalyst system variation on fatigue resistance is summarized in Figures 8 through 11. Analysis of ILD loss at both the first and third cycle leads to the following observations. Fatigue resistance is optimized:

polyol. Thus, the

manner

low levels

(including zero) of DBTDL,

~

at

~

at moderate to low levels of Niax A-12 at high levels of Niax A-433 at moderate to high levels of Dabco 33LV4

~

~

using combinations of either Niax catalyst with

~

Dabco 33LV.

This rather dramatic effect of the

fatigue was not expected unprecedented.

catalyst system on

Amine Amine

Figure

71. Catalyst Study 65% I LD Loss - 3

- Static Fatigue Cycle (%).

1 LD

catalyst (Union Carbide Corp.) catalyst (Union Carbide Corp.) Triethylenediamine, 33% solution in dipropylene ether glycol (Air Products and Chemicals Inc.) 3

4

10. Catalyst Study - Static Fatigue 25% I LD Loss - 3 Cycle (7~.

and its demonstration is

Figure 8. Catalyst Study - Static Fatigue 25% Loss - 1 Cycle (%). 2

Figure

An additional result of this study was the emergence of the adduct isocyanate as a powerful processing aid. In most systems, catalyst is used more as a processing aid than to control properties. In this system the processing latitude afforded by the

53

adduct isocyanate has presented the opportunity for study of catalyst primarily for its effect on properties. Indeed it has been shown that some of the catalyst and polyol combinations used in this study will not produce foam in other HR systems. As expected, catalysts also have a major effect on other foam properties. The principal effects are on hardness, tear strength, compression set, and percent change in C/D after humid aging. These effects are shown graphically in Figures 12 through 16 and are summarized as follows: 0 Hardness (Figures 12-13) - DBTDL has little or no effect on hardness as measured by ILD at 25 or 65% or by C/D at 50%. The Niax catalysts have a significant effect on hardness, both giving harder foams at low levels. In absolute terms, Niax A-4 gives softer foams than Niax A-1. Also, with Niax A-1 hardness and fatigue resistance are optimized at low levels, whereas with Niax A-4, optimization of these properties occurs at opposite extremes of the catalyst range studied. Dabco 33LV level has a less important effect, hardness increasing as level decreases.

o

Tear Strength (Figure 14) - DBTDL produces the largest effect, tear strength increasing as tin level increases. Niax A-1 produces the next largest effect, tear strength decreasing as Niax A-1 level decreases. The effect on tear strength of Niax A-4 and Dabco 33LV is less significant.

Figure 0

72.

Catalyst Study 25%

Catalyst Study

Tear

Strength iPLI ).

Compression Set (Figure 15) - Once again DBTDL produces a major effect, compression set decreasing as catalyst level decreases. For this property, Dabco 33LV level also has a significant effect, compression set decreasing as catalyst level increases. The Niax catalysts have little

Figure

14.

or no

effect.

1 LD (lobs1.

Figure 15. Catalyst Study Compression *

Figure 13. Catalyst Study 65% I LD (lobs) -

54

Set - 75%

(%).

Humid Aged CID, Yo Change (Figure 16) DBTDL and Dabco 33LV have little or no effect while the Niax catalysts have a significant effect, the percent change in Humid Aged C/D decreasing as Niax level is decreased. However, Niax A-1, at the level required for fatigue optimization, gives the minimum change in this property.

proves the quality of crosslinks, and therefore improves resistance to fatigue and compression set. Early competition between blowing and gelation (high DBTDL, low Dabco 33LV) promotes more random incorporation of urea hard segments, increases molecular weight and improves tear strength. While these statements regarding the role of catalyst seem

reasonable, they

are

admittedly speculative.

Considerable additional experimentation is required before the role of catalyst can be fully understood.

Other Variables

Using the optimum polyol (6500 M.W.) and catalyst combination of 0.2 parts Dabco 33LV and Figure 16. Catalyst Study Humid Aged C/D (% Change), 0

Other Prvperties --

Catalysts affect tensile

strength, elongation, and resilience, but the trends are less straightforward and the range of the effect is small. It is appropriate at this point to review the results of the catalyst phase of this study. ~ The wide processing latitude provided by the adduct isocyanate (HLG-3897) allowed a greater range of catalyst combinations to be studied. ~ Catalysts have a significant effect on fatigue and other HR foam properties. ~ A correlation exists between the effect of catalysts on fatigue and compression set, both properties responding similarily to catalyst changes. ~ On the other hand, those catalyst changes studied which improve fatigue resistance and compression set resistance are generally detrimental to tear strength. With respect to these last two observations, it is neither new nor surprising, but an example of a classic situation in the chemistry and physics of elastomeric polymers. Factors which emphasize or promote linearity, regularity and high molecular weight generally improve tear strength and elongation. Factors which increase or improve the nature of crosslinks improve compression set resistance. In this study there is a distinct parallel between compression set and fatigue resistance; hence, there is a strong indication that these two properties are influenced by some common factors in this foam system. Further analysis of these observations suggests that a relationship exists between catalyst function and property trends. High levels of DBTDL, primarily a catalyst for the urethane reaction (gelation), improve tear strength but worsen resistance to fatigue and compression set. The same effect is seen for low levels of Dabco 33LV, a strong catalyst for the urea reaction (blowing) as well as a gelation catalyst. These results suggest that predominance of the blowing reaction (high Dabco 33LV, low

DBTDL) early

in the sequence of events

formation of good hard segments

promotes

(polyureas),

im-

0.1 parts Niax A-1, the factors of water concentration, isocyanate index, and surfactant were evaluated for their effect on fatigue. None of these factors was found to adversely affect fatigue when tested in the normal range or concentrations used to prepare HR foams. Water level was checked between the concentrations of 2.4 and 3.0 phr. Index was varied between 96 and 108 while concentrations and types of surfactants considered to be practical in the preparation of HR foam were found to have no adverse influence. It is interesting to note that where less than the optimum polyol and catalyst combinations are used water level and index do have an effect on fatigue. Another iriteresting observation is that nominally identical polyols from different suppliers do not give the same fatigue properties.

Summary A detailed statistically designed study of fatigue HR foam made with adduct isocyanate (HLG-3897) has been carried out. Fatigue was monitored using a static fatigue test which correlates well with in-use testing. It has been shown that the polyol and the catalyst have a significant influence on the fatigue characteristics of this foam system. Polyol molecular weight, functionality, and chemical structure (e.g., percent primary hydroxyl) were found to be important in determining the effect of a given polyol system on fatigue. Within the scope of this study, a nominal 6500 M.W. primary triol gave the best fatigue resistance.

in

Catalyst had an equally important effect on fatigue. This phase of the study demonstrated that adduct isocyanate is an excellent processing aid allowing study of a wide range of catalyst systems, some of which would be insufficient for adequate processing in many HR foam systems. An optimum catalyst system for fatigue resistance consists of no more than 0.01 parts DBTDL, approximately 0.2 parts Dabco 33LV, and 0.1 parts Niax A-1. System variables such as water level, index and surfactant have only a minor effect on fatigue in a polyol and catalyst optimized system. continued

on

next page

55

J. D. Byam

H. W. Wolfe, Jr.

Harry W. Wolfe, Jr.,

re-

career

and Ph.D. degrees in Organic Chemistry from the University

on

a

of Delaware. Since joining the Du Pontt Company in 1954, he has had broad experience in both research and development in the field of urethane chemistry. Dr. Wolfe, is assigned to the Elastomer Chemicals Department at the Chestnut Run Laboratory. His current responsibilities involve both R & D and technical service for isocyanates. urethane polymers and flexible foam.

D. F. Brizzolara Donald F. Brizzolara received a B.S. in chemistry from Fordham University in 1963 and a Ph.D. in Organic Chemistry from New York University in 1968. He joined the R & D Division of the Du Pont

Elastomer Chemicals Departin 1968 as a research chemist where he had assignments in the area of fluorocarbon polymers. rubber latexes, and urethane chemistry. His current assignment is in the Sales Division. Chestnut Run Laboratory, where he has development and technical service responsibilities for molded and slab urethane foam, isocyanates, and urethane polymers. ment

56

John D. Byam’s industrial started with seven years in Research and Development, with the Plastics Department,

B.S. degree in Chemistry in 1951 from Lebanon Valley College, and the M.S. ceived

polymerization, rheology,

and physical test development. Plant process development and quality control of chemical and plastic operations occupied the next seven years. For the midOhio Valley Chapter of the ASQC. he taught practical applications of statistics to industrialists for several years. For over eleven years now, he has worked in Applied Mathematics for the Elastomers Department, where he leads a problem-solving seminar periodically conducted for the staff. Special areas of expertise are in the graphical design and analysis of experiments, the analysis of life or fatigue test data, elastomer rheology. process and regression analysis, and mathematical simulation of chemical and elastomer operations. He is author, either in whole or in part, of numeruus technical bulletilas. papers, and programs for computerizing statistical analyses. technical information recall and aanalysis, and simulation of the behavior of chemicals and elastomeric compounds in commercial operations.