Modeling and Simulation of the Creping Process

The manufacture of low density paper such as tissue and towel utilizes the creping process that consists of adhesively bonding the paper in wet state onto the surface of a smooth drying cylinder and scraping it off with a blade once the sheet is dry. In this paper, a fracture mechanics description of the creping process is presented. A mechanism of creping process is proposed as a periodic debonding with a mixed-mode fracture criterion with buckling of an elastic thin film. Finite element models validating creping mechanism will also be presented. Numerical calculations show results consistent with experimental data and known industrial observations. The model provides guidance in understanding and optimizing the creping process to produce high quality products. INTRODUCTION Creping process is the dominant micro-contraction process practiced in towel and tissue manufacturing. In this process, a continuous paper web is adhesively bonded to the surface of a large rotating drum (the Yankee Dryer), dried completely, and scraped off using a blade, typically at 5000 feet per minute. Prior to reaching the creping blade, the sheet is flat, smooth in most cases, and compact resembling the structure of writing paper. Immediately following the creping process, the sheet is "bulky" or has much lower density, soft or pliable, with low bending stiffness, and has regular surface undulations that cause the perception of softness, which is the primary end-use property of tissue grades. It is said, "tissue is made at the creping blade". Creping is a very poorly understood process, and the process operating parameters on a tissue machine have been primarily obtained through trial and error, and difficult to control. Yankee dryers are large steam-heated drums with polished cast iron surface. Yankee dryers can be as large as 24 feet in diameter and 18 feet in width. The surface speeds can be as high as 6500 feet per minute. The creping doctor blade is pressed against the drum to remove the adhesively bonded sheet. In the process, there is constant wear of the blade due to metal-to-metal contact and erosion of the Yankee dryer surface. The blade changes can be as frequent as every four hours resulting in lost production. Periodically, the Yankee dryer surface has to be reground to regain surface smoothness and uniformity. The steam heating systems has to be serviced periodically. In essence, the creping dryer is an enormously complex piece of equipment that is expensive to purchase and install and demanding to maintain. A systematic method to understand the creping process in a fundamental sense is needed in order to identify a generic set of fundamental process parameters such as energy input during debonding, the intensity or energy density required, and the best condition for the substrate when the energy is input to cause selective debonding, etc. Furthermore, the input parameters to the creping process that are significant should be identified based on published literature and industry knowledge and their effects on the sheet structure should be described in a computational model. BACKGROUND Hollmark [1] studied the creping process primarily through filming the process at high speed and developed a qualitative description of the creping process. He varied two parameters, namely, the creping angle, and the adhesion between the sheet and the dryer surface. The adhesion was changed by varying the percent resin content in the adhesive spray. Although it was concluded that the adhesion is probably the most important factor, specific adhesive properties were not identified. No attempt was made to develop analytical models for the process. Oliver [2] published his literature search on the creping process and made some general conclusions regarding the parameters that are important to the creping process, namely, creping blade angle, "adhesion", and the concentration of dissolved hemicellulose that affect the adhesion. Very few scientific publications exist in the area of tissue making. Almost all of them deal with the mechanism of adhesion of the paper web to the dryer surface, strictly through experimental observations. Any progress made in the creping area has been traditionally through PaperCon 2011 Page 1203 "unexpected results" described in numerous patents, without any description of the fundamentals of the process. There is no report of a systematic and coordinated analytical and experimental study of the creping process in the open literature. The pictures of the creping process, taken by Hollmark [1], have been used as the starting point to form a hypothesis for developing an analytical model [3]. The creping process can be schematically represented as shown in Figure 1 [3]. Figure 1. Mechanistic Description of Creping Process An idealized mechanics description of the creping process in its steady state is given below (see Figure 1): 1. Ahead of the creping blade, there exists a delamination length over which the sheet is not attached to the dryer. As the sheet is further pushed against the blade by the rotation of the dryer drum, crack propagation continues. 2. The crack continues to grow until it is energetically favorable for the free segment of the sheet to buckle. At this point the sheet buckles. 3. As the sheet is pushed further, some delamination growth takes place during buckling. 4. When the buckling is complete, the sheet completely collapses and forms the crepe fold. At this instance, the energy in the system drops well below the energy required for crack propagation, and the crack is arrested. 5. When the next segment of the sheet that is bonded to the dryer contacts the blade, delamination and buckling takes place. The entire cycle is repeated. From a mechanics point of view, the process is a continual competition between interfacial crack propagation and buckling. This concept has been analytically modeled by Sun [4]. CREPING EXPERIMENTS A laboratory creping simulator that can realistically simulate creping process up to 2500 feet per minute has been built at NC State University. The creping simulator has been used to study the process under a variety of conditions and results have been verified to correspond to expected large-scale trials [3].In this device, discrete paper samples can be attached to an electrically heated drum and scraped off with a creping blade mounted on a three axis load cell. The creping force which is the tangential direction to the drum is measured. In an experiment to understand the PaperCon 2011 Page 1204 creping process, a thin strip of paper, 2.5 in. wide and 6 in. long, with the longer dimension in the machine direction was bonded to the drum using Polyvinyl Alcohol (PVA) at 1.0% concentration. The paper was machine made and dried without creping. The basis weight of the paper used was 32 g/m. The Yankee dryer was run at 144 m/min (180rpm) and the creping angle was set to 80 degrees. The contact force of the creping blade and Yankee dryer in the normal direction was preset to 85N before creping. The tangential force trace was measured during creping process to study the effects of various process parameters. In order to understand the creping mechanism, the machine was slowed down to 15.8 m/min in order to capture the creping force during the process of creping, and a typical force trace is shown in Figure 2. Figure 2. Creping Force trace during creping process Each crepe fold formation is represented by each period in the trace from a low force to a peak and back to the low value. For example, the second crepe fold starts near 42 N, reaches a peak near 55 N, and when the crepe fold collapses, it reaches the low value again, and the cycle repeats with every crepe fold. The variation in the data is due to the variation in the paper structure since we are dealing with dimensions at the fiber length scale. It can be observed that the change in creping force in a single period is around 15 N, which is quite significant. CREPE MODELING AND NUMERICAL SIMULATION In this section, a finite element model for the creping process is presented in 2-D. The FEA model is generated by setting the creping problem using a crack propagation analysis available in ABAQUS/Standard. The Debonding Criterion ABAQUS allows three fracture criteria [5]: critical stress criterion, critical crack opening displacement (COD) criterion, and crack length versus time criterion. The critical stress criterion is more suitable to be applied in the creping problem to be studied here. The criterion is defined as