1. Introduction

1.1 Cleaning

Cleaning processes recently became more important, especially with regard to quality assurance. Cleanness is necessary to verify surfaces for compliance with quality or safety standards. The cleaning processes are needed in different life cycle phases of technical products [1].

For an efficient cleaning process, a number of requirements have to be met. Beside the contaminant that has to be removed, e.g. operating residues or chemical modified surface material, the degree of cleanliness as well as the measuring method has to be defined. Beyond this information, the application scenario is of high importance. Manufacturing, remanufacturing or recycling of used products represent a significant difference with regard to detailed product information. Furthermore, the infrastructure facilities are different.

1.2 Maintenance, repair and overhaul

Maintenance, repair and overhaul (MRO) plays a significant role for products with high investment costs and long life times such as aviation engines and gas turbines for electricity generation. Cleaning in the field of MRO is important for the inspection process of these products with regard to security relevant aspects. The information management of the operation conditions that influence security aspects is a challenge.

An on-site / on-wing inspection serves to check for critical failures, which do not in general affect the period of more detailed off-site checks. The inspection methods usually have to be suitable for a restricted accessibility. The dis- & re-assembly for a detailed off-site investigation are both time-consuming and expensive due to necessary equipment and man-power.

1.3 Carbon dioxide

Solid carbon dioxide (CO2) is a one-way blasting medium. Due to sublimation, no additional solid residues of the blasting medium remain beside the removed contaminant. CO2 is chemically inert. In solid state, the hardness of dry ice pellets is comparable with gypsum and the temperature at ambient pressure is -78.5°C. The CO2 used for blasting does not contribute to global warming, since it is either a by-product of the chemical industry or derived from natural sources [2].

1.4 Dry ice blasting, removal mechanisms

Cleaning with solid CO2 blasting is based on a combination of mechanical and thermal mechanisms, which is supported by the sublimation of the blasting media, Figure 1. Two forms of blasting with solid CO2 have to be distinguished: CO2-snow blasting and dry ice pellet blasting. Beyond this, the acceleration method, either by compressed air or by mechanical rotational wheel blasting, influences the energy efficiency as well as the blasting footprint.

2. Motivation

Expensive products like aviation engines or gas turbines for electricity generation have to be checked periodically. An investigation method for high cycle fatigue failure suitable for on-site restricted accessibility could help to save time and money of dis- and re-assembly. According to the results, the more detailed off-site check intervals could be adapted or even saved. Therefore, an application scenario was defined to search for a specific solution to clean and prepare the critical bottom failure of a running blade inside a turbine for inspection.

3. 90°-nozzle for CO2-snow blasting

The restricted accessibility of inner surfaces to be cleaned is usually not suitable for blasting processes, especially not for energy-efficient centrifugal blast cleaning. The already low efficiency of conventional compressed air-based blasting [3] becomes even lower. Therefore, a bent nozzle design for an ideal 90° blasting angle onto the surface with higher efficiency has to be developed.

3.1 Separation of functions

The mechanical effect is dependent on the orthogonal impetus of the particle on the surface, while its parallel proportion does not contribute. Because of this a 90° nozzle represent a solution for this basic problem. Regarding the design of the exemplarily presented CO2-snow nozzle, Figure 1, the generation of CO2-snow and the acceleration onto the surface to be cleaned will be separated. The orientation of the snow generation, which is investigated in another research project, will be changed. To save space, the expansion and the agglomeration chamber are designed parallel to the surface. Because of this, a bent tube is to be connected with the nozzle in a 90° turn.

Figure 2 shows three basic designs for this connection. Parameters are the diameter, the bending angle, and radius as well as process parameters, e. g. the mass flow rate of CO2, the pressure and the consumption of compressed air. These factors influence the agglomeration of snowflakes as well as the sublimation losses due to impacts on the surfaces while changing their direction. This 90° turn was designed regarding the experiences of the impact-sublimation relationship of investigations in rotational wheel blasting with dry ice pellets.

3.2 Energy efficiency

The efficiency of the generation of compressed air depends on the compressor as a combination of the pressure and the compressed air consumed. This efficiency was not part of the investigation, but can be combined with the compressor provider’s individual information. Therefore, the velocities of the particles as well as of the compressed air, Figure 2, were used to identify the design with the highest efficiency C, and closely followed up by A. The kinetic energy of the particles in relation to the consumption of compressed air represents the efficiency. In the following, the compressed air’s velocity and the respective distance of this velocity are used as a qualitative indicator for energy efficiency.

4. Simulation of acceleration

Based on the designs A to C, the blasting device was designed to investigate the compressed air consumption as shown in Figure 3. The modular setup was chosen to vary the dimensions and design of the various air supplies easily. First, investigation was carried out by computational fluid dynamics (CFD). Thus, an initial design and parameter were identified for both on-site application and the conventional off-site scenario. The verification of these results by investigations in the lab still has to be completed.

5. Results and discussion

Figure 4 shows exemplarily the comparison of a (110°-20°) turn (A) with a 90° turn (B). Furthermore, the designs have a different diameter reduction between the expansion chamber and the nozzle. The first (A) shows a continuous diameter reduction between the expansion chamber and the accelerating nozzle, the second (B) concentrates the reduction after the 90° turn in front of the nozzle. This results in different particle velocities as well as different flow characteristics inside the device.

The first design (A) shows a higher velocity of 294 m/s compared with 126 m/s of the second design (B) at the same process parameter setting. According to the investigations, this result is likely a consequence of the continuous diameter reduction. Thus, the acceleration of the nozzle becomes more effective.

The different contacts with the inner surface are likely a consequence of different bending characteristics. While the larger radius of B’s 90° bending results in a continuous contact of the particles, the smaller bending radiuses of A (110° turn combined with a 20° back turn) show concentrated contact areas.

Furthermore, the turning characteristics result in different footprints of the nozzles. As can be observed, A is more homogenous than B’s footprint, which shows two-centred particle distribution to the turning centre and the turning outside.

Figure 5 shows another two variants C and D of a 90° turn with additional orthogonal compressed air supply, which have only margin design differences. This supports turning the particles orientation onto the surface to be blasted as well as the CO2-particle’s acceleration. The comparison of C and D shows that the additional compressed air supply has a similar effect to the contact characteristic as the design shown in Figure 4: D is similar to B (disregarding the velocity), C shows concentrated contact areas comparable to A.

The energy efficiency of the device’s designs was compared to the kinetic energy rate of the CO2-snow particles in J/s divided by the compressed air consumption rate in g/s at a given blasting pressure. This results in an efficiency indicator in J/g, which is not suitable for comparing different pressure settings, but does not depend on the compressor’s efficiency. The efficiency, according to this indicator, showed a maximum of 5.5 kJ/kg and a minimum of 3.9 kJ/kg, representing a difference of 41%.

6. Conclusions and outlook

An on-site or on-wing cleaning and inspection method has a high potential to save time and costs of dis- and re-assembly. The limited accessibility requires the development of a specific blasting process, which offers an efficient acceleration method orthogonal onto the surface. Both the blasting device’s design and the process parameter settings have a significant influence on the particle’s velocity and kinetic energy as well as the flow characteristic.

Especially with regard to a non-durable one-way blasting media like solid CO2, the impacts on the inner surfaces are of interest. Whether these impacts both increase the agglomeration effect and generate larger snow particles which are more abrasive or lead to early sublimation losses still has to be investigated. Because of this multifunctional relationship, it is likely that instead of a universal design for each process parameter variation in a certain range, a specific device and nozzle design has to be developed.

The research project was funded by the Deutsche Forschungsgemeinschaft (DFG).

7. References

[1] Uhlmann, E.; Hollan, R.; Veit, R.; El Mernissi, A.: A Laser-Assisted Dry Ice Blasting Approach for Surface Cleaning. In: Proceedings of the 13th International Conference on LCE, Leuven, 2006
[2] Hollan, R. Uhlmann, E.: Energy-Efficient Cleaning and Pre-Treatment by Centrifugal Wheel Blasting with Sensitive Blasting Media. In: Proceedings of the 15th International Conference on LCE, Sydney, 2008
[3] Hollan, R. Uhlmann, E.: Sustainable Process Chains – Cleaning Technology and Energy Efficient Processes. In: Proceedings of the 16th International Conference on LCE, Cairo, 2009

Institute for Machine Tools and Factory
Management, Technical University Berlin
Pascalstraße 8-9, 10587 Berlin, Germany
Tel. +49.30.314 22413
Fax +49.30.314 25895
E-mail: hollan@iwf.tu-berlin.de
www.iwf.tu-berlin.de