Thrust and weight estimation for Doc. 29 noise models

Using ACMS data to more accurately predict noise levels at Amsterdam Airport Schiphol

Master Thesis (2026)
Author(s)

E.S.D. van Pijlen (TU Delft - Aerospace Engineering)

Contributor(s)

M. Snellen – Mentor (TU Delft - Control & Operations)

R.C. Van der Grift – Mentor (TU Delft - Operations & Environment)

J.A. Melkert – Graduation committee member (TU Delft - Flight Performance and Propulsion)

A. Amiri Simkooei – Graduation committee member (TU Delft - Operations & Environment)

J. Sun – Graduation committee member

More Info
expand_more
Publication Year
2026
Language
English
Graduation Date
28-01-2026
Awarding Institution
Programme
Aerospace Engineering, Flight Performance and Propulsion
Downloads counter
80
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

This paper investigates whether aircraft noise modeling can be improved by more accurately predicting the aircraft weight and thrust compared to the current methodology ECAC (European Civil Aviation Conference) described in Doc. 29 (Document 29). Using ACMS (Aircraft Condition Monitoring System) data from multiple aircraft types, two new weight estimation methods are proposed for departures: a climb slope and distance based approach, and a specific-energy method. The MAPE (Mean Average Percentage Error) of the current stage length approach is compared to the newly proposed methods. For thrust estimation, departures during the initial take-off roll and climb out are modeled using weight-dependent interpolations of the FPPs (Fixed-Point Profiles). For the other parts of the departure process, median FPPs, for which boundaries are determined by a flight segmentation model, are used. Arrival thrust values are predicted using a random forest regression model trained on flight path angle, calibrated airspeed, and corrected net thrust. This random forest model accurately captures thrust peak magnitudes and locations for most flights. Noise contour plots are generated for an original Doc. 29 model, an ACMS Doc. 29 and a new weight and thrust Doc. 29 model. For the ACMS Doc. 29 model, the ACMS thrust and weight data is directly used as input data for the noise model. The new weight and thrust estimates reveal closer agreement with ACMS Doc. 29 contours than with the original Doc. 29 method. This result indicates the rigidity of the FPPs and outdated ANP (Aircraft Noise Performance) database entries contribute to current modeling inaccuracies. The results demonstrate that the incorporation of performance relationships can significantly improve the theoretical Doc. 29 model.

Files

Thesis_Eric_van_Pijlen.pdf
(pdf | 28.6 Mb)
License info not available